sign in sign up
<<
Home , Cannabidiol

Drug Testing Research article and Analysis

Received: 6 June 2016 Revised: 9 July 2016 Accepted: 10 July 2016 Published online in Wiley Online Library: 10 August 2016

(www.drugtestinganalysis.com) DOI 10.1002/dta.2042 Development and validation of an ultra-fast and sensitive microflow liquid chromatography- tandem mass spectrometry (MFLC-MS/MS) method for quantification of LSD and its metabolites in plasma and application to a controlled LSD administration study in humans Andrea E. Steuer,a* Michael Poetzsch,a Lorena Stock,a Lisa Eisenbeiss,a Yasmin Schmid,b Matthias E. Liechtib and Thomas Kraemera

Lysergic acid diethylamide (LSD) is a semi-synthetic that has gained popularity as a recreational drug and has been investigated as an adjunct to psychotherapy. Analysis of LSD represents a major challenge in forensic toxicology due to its insta- bility, low drug concentrations, and short detection windows in biological samples. A new, fast, and sensitive microflow liquid chromatography (MFLC) tandem mass spectrometry method for the validated quantification of LSD, iso-LSD, 2-oxo 3-hydroxy- LSD (oxo-HO-LSD), and N-desmethyl-LSD (nor-LSD) was developed in plasma and applied to a controlled pharmacokinetic (PK) study in humans to test whether LSD metabolites would offer for longer detection windows. Five hundred microlitres of plasma were extracted by solid phase extraction. Analysis was performed on a Sciex Eksigent MFLC system coupled to a Sciex 5500 QTrap. The method was validated according to (inter)-national guidelines. MFLC allowed for separation of the mentioned analytes within 3 minutes and limits of quantification of 0.01 ng/mL. Validation criteria were fulfilled for all analytes. PK data could be calculated for LSD, iso-LSD, and oxo-HO-LSD in all participants. Additionally, hydroxy-LSD (HO-LSD) and HO-LSD glucuronide could be qual- itatively detected and PK determined in 11 and 8 subjects, respectively. Nor-LSD was only sporadically detected. Elimination half- lives of iso-LSD (median 12 h) and LSD metabolites (median 9, 7.4, 12, and 11 h for oxo-HO-LSD, HO-LSD, HO-LSD-gluc, and nor- LSD, respectively) exceeded those of LSD (median 4.2 h). However, screening for metabolites to increase detection windows in plasma seems not to be constructive due to their very low concentrations. Copyright © 2016 John Wiley & Sons, Ltd.

Keywords: microflow LC-MS/MS; LSD; LSD metabolites;

Introduction LSD consists of an skeleton in which the (5R,8R)-config- uration confers psychoactive properties. Iso-LSD is a diastereomer Lysergic acid diethylamide (LSD) is a semi-synthetic classic seroto- of LSD which was shown to be formed during the production of nergic hallucinogen derived from lysergic acid, a natural substance LSD and/or under basic conditions[9,10] and is therefore an addi- from the parasitic rye fungus Claviceps purpurea. During the 1950s, tional marker for LSD consumption.[10] The main metabolites of LSD was introduced to the medical community as an experimental LSD described in urine as further potentially ingestions markers tool to induce temporary psychotic-like states in healthy volunteers are 2-oxo-3-hydroxy-LSD (oxo-HO-LSD) and N-desmethyl-LSD (model-) and later to enhance psychotherapeutic treat- (nor-LSD).[11] Further minor metabolites detected in human urine ments (psycholytic or psychedelic therapy).[1,2] In the 1960s, it gained popularity as a recreational drug and it is still one of the most often used hallucinogenic substances worldwide.[1,3] As one * Correspondence to: Andrea E. Steuer, Department of Forensic Pharmacology and of the most potent , recreational LSD use in doses Toxicology, University of Zurich, Zurich Institute of Forensic (ZIFM), μ Winterthurerstrasse 190/52, CH-8057 Zurich, Switzerland. as low as 25 to 200 g are associated with significant alterations E-mail: [email protected] in consciousness. Under controlled conditions in clinical studies, – LSD is generally well tolerated.[1,4 6] Additionally, LSD has been a Department of Forensic Pharmacology and Toxicology, Zurich Institute of shown to acutely alter emotion processing in ways that may be use- Forensic Medicine, University of Zurich, Switzerland [6] ful for LSD-assisted psychotherapy. However, under uncontrolled b Psychopharmacology Research, Division of Clinical Pharmacology and 788 conditions, adverse psychological reactions including panic are Toxicology, Department of Biomedicine and Department of , common effects of LSD use.[7,8] University Hospital Basel, Basel, Switzerland Drug Testing New microflow LC-MS/MS method for quantification of LSD and its metabolites and application and Analysis

so far were nor-iso-LSD, N-deethyl-LSD/lysergic acid ethylamide deuterated internal standard (IS) LSD-d3 (purity > 99%) were ob- (LAE), trioxylated-LSD, di-hydroxy-LSD (di-HO-LSD), lysergic acid tained from Cerilliant (delivered by Sigma-Aldrich, Buchs, ethyl-2-hydroxyethylamide (LEO), and glucuronides of 13- and 14- Switzerland). Nor-LSD (purity > 97%) was obtained as a solid from hydroxy-LSD (HO-LSD).[11,12] Analysis of LSD represents a major Toronto Research Chemicals (Toronto, Canada). Isolute HCX car- challenge in clinical and forensic toxicology especially due to its tridges (130 mg, 3 mL) were obtained from Biotage (Uppsala, instability[13–15] and rather low drug and metabolite concentrations Sweden). Water was purified with a Purelab Ultra millipore filtration in biological samples.[1,16,17] unit (Labtech, Villmergen, Switzerland) and acetonitrile of HPLC Immunoassay prescreens are typically performed in clinical and grade was obtained from Fluka (Buchs, Switzerland). All other forensic toxicology to allow for fast differentiation between negative chemicals used were from Merck (Zug, Switzerland) and of the and positive presumptive tests. However, in terms of LSD, major highest grade available. cross-reactivity is known through or different fentanyl- [18] analogues. Applicationoffentanyliscommonasaninitialemer- Biosamples gency treatment, especially following traffic accidents. Consequently, high numbers of false-positive LSD immunoassays routinely occur. Human blank plasma samples were used for development and val- More sophisticated, confirmatory analysis on LSD detection and idation of the procedure and were obtained from healthy volun- quantification is mainly performed by gas chromatography-mass teers, collected as ammonium heparin blood, directly centrifuged spectrometry (GC-MS)[19,20] or liquid chromatography (LC)-MS(/MS) (5000 g, 15 min) and stored in aliquots at -20 ° C. techniques.[10,11,17,21–24] With these techniques, mainly LSD, iso-LSD, Plasma samples after controlled administration of LSD were col- oxo-HO-LSD, and rarely nor-LSD have been detected in urine. In lected at the University Hospital of Basel within a controlled LSD ap- [4,16] blood, LSD was the only target with rather short detection plication study to humans. Samples were stored protected windows.[25] In forensics, larger detection windows are desirable or from light at -80 °C till analysis (max. 18 months) as described in de- even necessary, especially if urine collection is impossible. Additional tail below. detection of metabolites also in blood might be an alternative but have not been investigated in detail so far. Oxo-HO-LSD was only Initial screening for LSD metabolites detected in postmortem blood[17] and recently in several emergency toxicology cases.[21] Dolder et al. were the first to determine LSD and Five plasma samples (t = 2.5 h) were initially screened for the pres- ence of the following LSD metabolites: oxo-HO-LSD, nor-LSD, HO- particularly oxo-HO-LSD pharmacokinetics (PK) in plasma after a con- [11,12] trolled 200 μg of LSD using a rapid turbo-flow LC-MS/MS LSD, HO-LSD glucuronide, di-HO-LSD, LAE, and LEO. Two hun- μ method.[16,21] However, despite several advantages, the method dred L of plasma were protein precipitated by the addition of μ lacked sufficient sensitivity for quantification of oxo-HO-LSD, and 600 L acetonitrile, shaken for 10 min, centrifuged (10 000 g, μ PK calculations were only possible for one-half of the participants. 10 min) and 500 L supernatant were evaporated to dryness under μ Application of newer chromatographic strategies such as nano- a gentle stream of nitrogen at 40 °C. After reconstitution in 50 L μ or microflow LC (MFLC) might offer higher sensitivity for LSD when mobile phase, 5 L were injected into an ABSciex Eksigent coupled to electrospray ionization (ESI) MS. For instance, MFLC has Microflow LC system (Redwood City, CA, USA) coupled to a Sciex been successfully applied for quantification of tetrahydrocannabi- 5500 QTrap linear ion trap quadrupole mass spectrometer (Sciex, nol (THC) in very low concentration.[26,27] In general, scaling down Darmstadt/Germany) as described below. flow rates to nano LC or MFLC systems led to certain benefits such The MFLC settings were as follows: Halo® Phenyl Hexyl column μ as reduced solvent consumption, higher throughput through de- (50 x 0.5 mm, 2.7 m), gradient elution with 10 mM ammonium for- creased run times, a higher ionization yield, and reduced ion mate buffer in water pH 3.5 (A) and acetonitrile containing 0.1% μ suppression/enhancement effects.[28,29] A previous study demon- (v/v) formic acid (B). The flow rate was 30 L/min with the following – – strated similar performance characteristics in typical validation pa- gradient: start conditions 10% B; 0 2.1 min increase to 50% B; 2.1 rameters for MFLC compared to conventional LC.[30] However, 2.23 min to 80% B, hold until 2.6 min; re-equilibrating to start condi- despite its potential benefits, in the field of clinical and forensic tox- tions of 10%B and hold for 0.3 min. Total run time was 3 min. icology such methods are currently scarce, possibly because these The MS was operated in the enhanced product ion (EPI) scan – techniques are suspected to lack ruggedness.[28–30] mode using the following settings: mass range 50 1000, scan rate The aim of the present study was the development and valida- 10 000 Da/s, dynamic trap fill time. EPI scans were recorded for tion of a new, fast, and sensitive MFLC-MS/MS method for the vali- the expected protonated molecular ions of the potentially metabo- dated quantification of LSD, iso-LSD, oxo-HO-LSD, and nor-LSD in lites applying a collision energy spread (CES) of 35 +/- 15 eV. plasma. Additionally, other non-commercial LSD metabolites (HO- LSD, HO-LSD glucuronide, di-HO-LSD, LAE, and LEO) were screened Quantitative method for. The method was then applied to the analysis of human plasma Sample preparation samples from a controlled LSD administration study in 16 healthy volunteers[4] to test whether screening for metabolites would allow The sample preparation process was performed protected from for longer detection of LSD consumption in plasma relevant to light through the application of aluminum foil and light protected clinical and forensic toxicology. brown glass autosampler vials. Plasma samples were extracted by solid phase extraction (SPE) using HCX cartridges (130 mg, 3 mL). Briefly, to 500 μLplasma,50μLoftheIS(LSD-d3, 1 ng/mL) and Materials and methods 50 μL spiking (calibrator or quality control (QC)) or 50 μL acetonitrile were added and the mixture was thoroughly vortexed. Chemicals and reagents Two millilitres of water were added, vortexed and transferred to

Acetonitrilic (0.1 mg/mL) of LSD (purity > 99%), iso-LSD SPE cartridges, previously conditioned with 1 mL of methanol and 789 (purity > 99%), and oxo-HO-LSD (purity > 88%) and of the 1 mL of water. After passage of the supernatant, the cartridge was Drug Testing and Analysis A. E. Steuer et al.

washed with 1 mL of water, 1 mL of a 0.01 M HCl solution and 1 mL Selectivity of methanol. Elution was performed with 1 mL freshly prepared Ten blank plasma samples from different sources were analyzed for mixture of methanol/aqueous ammonia (98:2, v/v)intosilylated peaks interfering with the detection of analytes or IS. Two zero sam- brown glass autosampler vials. After neutralization of the mixture ples (blank sample + IS) were analyzed to check for appropriate IS with 100 μL of a 20% formic acid solution, the mixture was evapo- purity and presence of native analytes. rated to dryness under a gentle stream of nitrogen at 40 °C and reconstituted in 50 μL mobile phase A. The reconstituted samples Stability were filtrated with centrifugal units (VWR centrifugal filter, 0.45 μM Processed sample (benchtop) stability was investigated at QC LOW pore size, modified nylon membrane, VWR, Dietikon, Switzerland). [32] Aliquots of 5 μL were injected into the MFLC-MS/MS system. and HIGH concentration (n = 6 each) according to Peters. Limits LC-MS/MS settings The lowest point of the calibration curve was defined as the limit of The analysis was performed using a Sciex Eksigent Microflow LC quantification (LOQ) of the method and had to meet the criteria of system (Redwood City, CA, USA) coupled to a Sciex 5500 QTrap signal to noise ratio of 10:1 determined via peak heights, impreci- linear ion trap quadrupole mass spectrometer (Sciex, sion lower than 20%, and accuracy between 80 and120%. Darmstadt/Germany). The MFLC settings were as follows: Halo® Phenyl Hexyl column Calibration model (50 x 0.5 mm, 2.7 μm), gradient elution with 10 mM ammonium for- Daily calibration curves (single measurement per level) were pre- mate buffer in water pH 3.5 (A) and acetonitrile containing 0.1% pared with each batch of validation samples. Calibration model (v/v) formic acid (B). The flow rate was 30 μL/min with the following was evaluated based on the bias and imprecision data. The regres- gradient: start conditions 10% B; 0–2.1 min increase to 50% B; sion lines were calculated using a linear weighted 1/X and 1/X2 2.1–2.23 min to 80% B, hold until 2.6 min; re-equilibrating to start least-squares regression model. The back-calculated concentrations conditions of 10%B and hold for 0.3 min. Total run time was 3 min. of all calibration samples were compared to their respective nomi- The Turbo V ion source equipped with a hybrid electrode (50 μm nal values and quantitative accuracy was required within 20% of internal diameter) was operated in positive ESI mode with the fol- target. lowing MS conditions: gas 1, nitrogen (30 psi); gas 2, nitrogen (35 psi); ion spray voltage, 5500 V; ion-source temperature, 150 °C; Recovery and matrix effects curtain gas, nitrogen (20 psi), collision gas, medium. The MS was op- Recovery (RE) and matrix effect (ME) were evaluated at QC LOW and erated in multiple reaction monitoring (MRM) mode using 2 to 3 HIGH concentration using 6 different plasma sources according to transitions for each analyte except for the IS where 1 MRM transi- the simplified approach described by Matuszewski et al.[33] tion was applied. The MS settings were as follows: LSD 324/223, 324/208, 324/207; oxo-HO-LSD 356/237, 356/222, 356/139; nor- Bias, precision, and beta tolerance interval LSD 310/209, 310/74, 310/237; iso-LSD 324/208, 324/223, 324/207; QC samples (LOW, MED, HIGH) were analyzed according to the pro- LSD-d3 327/226. For secure analyte identification both quantifier and qualifier ions had to be present at the same retention time as cedures described above in duplicate on each of eight days. Bias the analyte in the QC sample and the quantifier qualifier ratio had was calculated as the percent deviation of the mean calculated con- to be within the acceptable range (20–50% depending on the rela- centration at each concentration level from the corresponding the- tive intensity to the quantifier) of those obtained from the QC oretical concentration. Intra-day (RSDR) and inter-day (RSDT) [31] imprecision were calculated as relative standard deviation (RSD) ac- samples. In a second method, the following MRM transitions [32] were additionally included for screening and relative quantification cording to Peters. The lower (Ll) and upper (Lu) limit of the 95% beta tolerance interval were calculated using the following equa- of HO-LSD 340/239, 340/297 and HO-LSD glucuronide 516/340, [32] 516/239. Quantifiers are underlined, respectively. The MS was con- tions simplified for duplicate measurements on eight days : trolled by analyst 1.6.2 software. 1Ll [%] = Bias [%] – 2.508 * RSDT [%] 2Lu [%] = Bias [%] + 2.508 * RSDT [%] Method validation Applicability/LSD and metabolites pharmacokinetics Preparation of calibration and QC samples LSD study Separate commercially available 0.1 mg/mL acetonitrilic solutions for each analyte were used as stock solutions for calibration and The developed method has been applied to the analysis of plasma QC spiking solutions. Different working solutions (10–10 samples collected in a controlled LSD administration study to 000 ng/mL) were prepared by dilution from each stock solution in humans as described in detail elsewhere.[4,16] Briefly, LSD hydrate acetonitrile. Calibration standards and QC samples were prepared with a purity (HPLC) of 99.7% was obtained from Lipomed AG, by mixing appropriate amounts of the corresponding stock or Arlesheim, Switzerland, and prepared in gelatine capsules. The cap- working solution to give the final calibration spiking solutions (Cal sules were stored in closed containers kept dry and in the dark at 1–7) and QC spiking solution (LOW, MED, HIGH) in a final concentra- 4 °C until use within 1–7 months after production. Identity and pu- tion 10 times higher than the corresponding plasma concentra- rity was again confirmed within the capsules after the study. tions. Calibration standards and QC samples (LOW, MED, HIGH) Placebo and a single absolute dose of 200 μg LSD hydrate, corre- were prepared from 500 μL analyte-free plasma and 50 μLofthe sponding to a dose of 2.84 ± 0.5 μg/ kg body weight (mean ± SD; corresponding fortifying solution. The final plasma calibrator con- range: 2.04–3.85 μg) were administered orally without food to 16

790 centrations were 0.01, 0.1, 0.5, 1, 5, 10, 20 ng/mL and QC concentra- healthy subjects (8 men, 8 women) in a double-blind, placebo- tions were 0.02, 0.75, 15 ng/mL, respectively for each analyte. controlled, randomized crossover design. The washout periods Drug Testing New microflow LC-MS/MS method for quantification of LSD and its metabolites and application and Analysis between sessions were at least 7 days. Blood samples were col- could not be reached even with optimized conditions and pure lected into lithium heparin tubes 1 hour before and 0.5, 1, 1.5, 2.5, compound solutions. Switching to a Sciex5500 Qtrap which shares 3, 4, 6, 8, 10, 12, 16, and 24 h after LSD administration. The samples the same geometry and ion source, but provides a larger ion inlet were immediately centrifuged and the plasma rapidly stored at - diameter increased peak areas by a factor of approximately 10 20 °C until analysis within 18 months. The study was conducted in and allowed detection of concentrations of 0.01 ng/mL in solution. accordance with the Declaration of Helsinki and International Con- ference on Harmonization Guidelines in Good Clinical Practice and Chromatography approved by the Ethics Committee of the Canton of Basel, Chromatography was performed using a Phenyl-Hexyl column Switzerland and the Swiss Agency for Therapeutic Products which allowed for baseline separation of oxo-HO-LSD, LSD and (Swissmedic). The administration of LSD to healthy subjects was au- iso-LSD within 3 min (including re-equilibration) as shown in thorized by the Swiss Federal Office for Public Health, Bern, Figure 1. Nor-LSD only slightly overlapped with LSD (resolution Switzerland. The study was registered at ClinicalTrials.gov R = 0.7). Different gradient conditions, solvents and stationary (NCT01878942). All of the subjects provided written informed con- phase (C18) were tested, but did not result in improved resolution. sent after being given written and oral descriptions of the study, the However, the diversity of commercially available stationary phases procedures involved, and the effects and possible risks of LSD for MFLC is low compared to conventional LC columns. In addition, administration. high buffer concentrations are known to cause clogging of the Plasma pharmacokinetics MFLC system, thus limiting reasonable variations of chromato- graphic settings. No cross-talk could be observed between LSD, Pharmacokinetic calculations (maximum plasma concentration LSD-d3 and nor-LSD. Therefore, the slight overlap was accepted in (Cmax), time to reach maximum plasma concentration (tmax), area terms of throughput and robustness of the method. In a previous – under the plasma concentration-time curve from 0 24 h (AUC24), study on 40 and neuroleptics,[30] the MFLC system AUC from time 0 to infinity (AUC∞), elimination half-life (t1/2), clear- was prone to retention time shifts, especially with varying analyte ance (Cl/F), volume of distribution (Vd/F) were performed using concentrations. The same effect could not be observed in the cur- non-compartmental methods with PK solutions 2.0 software, Sum- rent method on LSD and metabolites with stable retention times mit Research Services (Montrose, CO, USA). (+/- 0.02 min) observed over a time period of approximately four months and over the whole calibration range. Nevertheless, to Results and discussion avoid unnecessary reanalysis of samples due to partly shifting re- tention times, it was initially decided to perform analysis in MRM The aim of the presented study was the development and valida- mode without pre-scheduled time windows. As only four analytes tion of a sensitive method for quantification of LSD, iso-LSD and with three transitions each plus one IS had to be monitored in the its metabolites oxo-HO-LSD and nor-LSD using MFLC-MS/MS. Addi- quantitative method, still more than 10 data points, commonly rec- tionally, HO-LSD and HO-LSD glucuronide were included for ommended for quantitative purposes, could be gathered for each qualitative/semi-quantitative determination although not available peak even in the lowest concentration. as a reference material. The presented method was then applied to the analysis of samples from a controlled LSD administration study Sample preparation in humans and provided for the first time sensitive PK information Sample preparation can be performed using a variety of different for oxo-HO-LSD and iso-LSD as well as the detection of HO-LSD procedures. Previous methods for LSD successfully applied liquid- and its glucuronide in human plasma samples. liquid extraction (LLE),[10,11,17,22] SPE,[20,23] and online extraction.[21] Simple, fast, and cost-effective protein precipitation was suspected Method development in an earlier study to cause instability of the MFLC system[30] and was therefore omitted in the current method for larger sample se- MS detection ries. LLE was considered a valuable alternative as it has been a very MRM transitions and their particular settings, collision energy (CE) common sample workup procedure in forensic toxicology due to and potentials were specific for each analyte and were determined characteristics such as being rapid, simple, cheap, and suitable for using an acetonitrilic/aqueous solution (0.01 mg/mL) of each ana- a broad range of substances.[34] Therefore, different LLE procedures lyte injected by the integrated syringe pump and obtained using were tested: (a) butyl acetate, (b) butyl acetate/ethyl acetate (1:1), the automatic compound optimization function on the Qtrap (c) diethyl ether/ethyl acetate (1:1), and 2-propanol/dichlorometh- 5500. The most abundant MRM transition was used as quantifier, ane/ethyl acetate (1:1:1). While LLE at pH 9 with butyl acetate pro- the others as qualifiers. The chosen MRM transitions allowed for un- vided the best results with recoveries above 50% for LSD, iso-LSD ambiguous identification and differentiation of LSD and its metab- and nor-LSD, recoveries were only 10% for oxo-HO-LSD (data not olites oxo-HO-LSD and nor-LSD. As expected, LSD and iso-LSD as shown), most likely due to its higher hydrophilicity. The lowest cal- diastereomers shared the same transitions, however in different ibrator of 0.01 ng/mL was undetectable after LLE. These findings ion intensities resulting in m/z 324/223 as the highest abundant generally match those of others studies.[17,22] Switching to SPE with transition for LSD and m/z 324/208 for iso-LSD, respectively. Addi- cation exchange sorbents like HCX finally allowed the detection of tionally, to account for different flow settings on an MFLC device all analytes including oxo-HO-LSD in the desired low concentration compared to conventional chromatographic systems, declustering range. SPE can be automated if higher throughput is desired. As al- potential, entrance potential, source temperature and gas settings ready observed in earlier studies, MFLC still bears some drawbacks, were optimized by flow injection experiments to obtain maximum the most severe being clogging of MFLC columns and ESI sensitivity. Initial measurements were performed using the MFLC capillary.[30] Occurrence of these problems appeared randomly system coupled to a Sciex 4500 Qtrap as the routine configuration and could not be correlated with a certain number of injections. 791 in our lab. However, target concentrations of at least 0.01 ng/mL Even with the selected more sophisticated sample preparation, that Drug Testing and Analysis A. E. Steuer et al.

Figure 1. Microflow LC-MS/MS extracted ion chromatograms of a plasma SPE containing oxo-HO-LSD (MRM 356/237), nor-LSD (MRM 310/209), LSD (324/223, 208) and iso-LSD (MRM 324/223, 208) at 0.5 ng/mL each. MRM, multiple reaction monitoring.

problem could only be circumvented by introduction of an addi- method. However, considering that MFLC was suspected to be tional filtration step prior to injection. Unspecific binding of LSD, less susceptible to ME 28, 29 the observed effects were rather sur- iso-LSD and its metabolites to the filtration membrane was tested prising. Causes other than matrix interferences could have been and excluded by comparing peak areas prior to and after filtration responsible for the signal reduction, for example loss of LSD and (data not shown). metabolites during the evaporation step. LSD is known to un- dergo conversion to iso-LSD under high temperature and alka- [9,15] Method validation line conditions. However, matrix effects observed for LSD were in the same range as for iso-LSD with no indication for in- The described procedure was validated according to recommenda- creased iso-LSD concentrations during the extraction process. tions on method validation in the context of quality management Furthermore, matrix-free solutions were additionally prepared af- with forensic-toxicological investigations published by the Society ter evaporation out of MeOH/NH solution as performed during [32] 3 of Toxicological and Forensic Chemistry (GTFCh) and interna- the SPE clean-up, but no significant differences in peak areas [35–39] tionally accepted recommendations. were observed. Selectivity Linearity and limits of detection No interfering peaks from matrix or the IS solution were detected proving sufficient selectivity by the chosen MRM transitions. Calibration ranges were chosen with a special focus of the very low concentration range aiming to accurately and precisely detect at Recovery and matrix effects least 0.01 ng/mL of all analytes. Emphasis of the calibration range RE and ME data are listed in Table 1. LSD, iso-LSD, and nor-LSD was set to lower concentrations which should cover the majority could be extracted with REs of approximately 100%. Coefficients of real LSD cases, at least in a controlled pharmacokinetic study. Ad- of variations (CVs) from six different plasma samples were ditionally, to handle potential cases of intended or accidental over- slightly above the acceptance criteria of +/- 15% (20% at QC doses – where blood concentrations up to 14 ng/mL and higher LOW). Inclusion of the IS as is customary for quantification gave were observed[40] – the calibration range was extended on the up- acceptable CVs of 10% or below. RE of oxo-HO-LSD was 73% per end to cover levels up to 20 ng/mL. Calibration curves using and 87% for QC LOW and HIGH, respectively with acceptable seven concentration levels with six replicates each were con- CVs and proved to be sufficient for sensitive determination of structed to evaluate the calibration model. All analytes showed lin- oxo-HO-LSD. Matrix effects calculated from six different plasma earity over the selected calibration range. A weighted 1/X2 sources were about 70% for all analytes with CVs within the ac- calibration model was used to account for unequal variances ceptance criteria. Sensitivity was still sufficient, although signal (heteroscedasticity) across the calibration range. For the accuracy suppression of approximately 30% was observed. Taking into ac- and precision experiments, daily calibration curves were prepared

792 count the high reproducibility between different plasma sources with each batch of QC samples. The back-calculated concentrations the detected suppression could be tolerated for the presented of all calibration samples were +/- 20% of target concentration. Drug Testing New microflow LC-MS/MS method for quantification of LSD and its metabolites and application and Analysis

Table 1. Summary of validation parameters for LSD and its metabolites. Values outside the acceptance criteria are given in italic print. RE, recovery; ME, matrix effects; IS corr., corrected by internal standard LSD-d3, RSD, relative standard deviation; RSDR, intraday precision; RSDT, inter-day precision; Ll lower limit of the 95% beta tolerance interval; Lu upper limit of the 95% beta tolerance interval

Analyte Conc. ng/mL RE % (% RSD) RE IS corr. % (% RSD) ME % (% RSD) ME IS corr. % (% RSD) Bias % RSDR %RSDT %Ll %Lu %

LSD QC low 0.02 91.8 (16.4) 93.9 (8.7) 68.6 (10.5) 89.7 (11.7) -0.2 6.8 9.9 -25.1 24.8 QC med 0.75 -4.3 5.3 5.6 -18.3 9.6 QC high 15 104.6 (17.8) 100.0 (2.6) 66.9 (11.5) 112.0 (5.5) -2.9 3.4 3.5 -11.7 6.0 Iso-LSD QC low 0.02 104.9 (23.3) 106.5 (10.5) 66.8 (17.0) 88.4 (8.5) -11.0 13.2 20.9 -63.441.4 QC med 0.75 -3.1 3.9 10.7 -29.8 23.7 QC high 15 106.6 (26.3) 100.9 (7.9) 66.5 (10.1) 110.3 (3.5) -7.0 6.1 9.0 -29.6 15.6 Oxo-HO-LSD QC low 0.02 73.4 (19.4) 74.4 (25.2) 72.9 (15.8) 98.5 (16.7) 10.6 9.8 17.0 -32.0 53.3 QC med 0.75 . 0.4 13.2 12.9 -32.1 32.9 QC high 15 88.6 (8.0) 81.6 (14.4) 65.5 (10.1) 109.0 (8.8) -12.8 11.3 14.7 -49.6 23.9 Nor-LSD QC low 0.02 101.0 (21.8) 102.8 (10.9) 72.7 (11.7) 96.5 (5.8) 10.2 5.7 10.9 -17.2 37.5 QC med 0.75 5.5 5.9 4.5 -5.7 16.8 QC high 15 107.4 (13.6) 103.4 (8.9) 62.4 (12.2) 104.4 (5.8) -5.4 4.5 5.3 -18.7 7.8

LOQs corresponded to the lowest point of the calibration curve Stability (0.01 ng/mL) for all analytes, fulfilling the criteria of S/N greater than In general, LSD in solution is known for its instability, mainly at ele- 10 and residuals within +/- 20%. The combination of SPE with MFLC vated temperatures, under alkaline conditions, sun or ultraviolet allowed LOQs of 0.01 ng/mL for all analytes. Considering the light.[13–15] Long-term stability (6 months) has been shown for sample volume used, the LOQs in the current method were consid- LSD, oxo-hydroxy-LSD, and nor-LSD when kept under refrigerated erably lower than in previous studies, particularly for oxo- or frozen conditions.[14,23] Analysis of the current PK samples was HO-LSD.[23] In general, other studies required higher sample performed from plasma samples immediately frozen at -80 °C for amounts to reach comparable LOQs for LSD, iso-LSD, and nor- a maximum of 18 months and thawed the first time for the analysis LSD.[10,20,22,23] For instance, Chung et al. reached similar LOQs for in the current study. Therefore, further long-term stability experi- LSD and oxo-HO-LSD but used double as much blood sample,[17] ments were not performed during the method validation process. Favretto et al. even needed 2 mL of blood to achieve an LOQ of In post-processed samples, the analytes were shown to be stable 0.02 ng/mL for the mentioned analytes.[22] for a period of at least 35 h at 4 °C protected from light (autosampler Accuracy, precision, and beta tolerance intervals conditions). Especially no decrease of LSD and increase of iso-LSD could be observed indicating no relevant conversion of LSD to QC samples (LOW, MED, and HIGH) were analyzed in duplicate on iso-LSD during the sample analysis. each of eight days as it has been proposed by Peters et al.[32] The concentrations in the QC samples were calculated based on the daily calibration curves. Accuracy, repeatability, and time-different Applicability/pharmacokinetic analysis of LSD and metabolites intermediate precision were calculated as described above. The re- Initial screening for LSD metabolites sults obtained using calibration curves are shown in Table 2. The re- sults above the mentioned acceptance criteria are marked in italic. To date, only few data on human LSD exist.[11,12,41] All analytes fulfilled the specific validation parameters. Additionally, Oxo-HO-LSD in urine[11,42] and according to recent studies in beta tolerance intervals were calculated. The beta tolerance interval blood/plasma[16,21] is considered the main LSD metabolite and rep- describes the range that includes an actual measured concentra- resents the target analyte for identification of LSD consumption tion with a probability of 95%. Except for iso-LSD in QC LOW and next to LSD itself. To the best of our knowledge, detection of other oxo-HO-LSD in QC HIGH also the beta tolerance intervals fulfilled presumably minor metabolites in human urine was reported in only the recommended criteria of +/- 30% (40% for the lowest one study,[11] and except for nor-LSD these are typically not rou- concentrations). tinely screened for. Only oxo-HO-LSD and nor-LSD are

Table 2. Summary of pharmacokinetic parameters of LSD and its metabolites. All parameters are given as median and range of N participants

Analyte N Cmax,ng/mL tmax,h t1/2,h AUC24, AUC∞, Cl/F Vd/F ng/mL*h-1 ng/mL*h-1 (mL/hr/kg) (mL/kg)

LSD 16 3.5 (2.0-7.4) 1.5 (0.50-5.0) 4.2 (2.0-5.6) 21 (11-40) 22 (11-41) 140 (71-220) 660 (460-1300) Iso-LSD 16 0.70 (0.43-1.7) 2.3 (0.50-6.0) 12 (8.4-24) 9.2 (5.4-19) 16 (7.4-26) Oxo-HO-LSD 16 0.11 (0.05-0.23) 4.0 (3.0-16) 9.0 (4.2-18) 1.2 (0.67-2.6) 1.6 (0.82-3.8) HO-LSD* 11 0.03 (0.02-0.12) 2.0 (1.5-4.0) 7.4 (5.0-17) 0.24 (0.13-1.3) 0.31 (0.14-1.5) HO-LSD gluc* 8 0.02 (0.01-0.04) 3.0 (1.5-6.0) 12 (6.0-17) 0.24 (0.09-0.52) 0.33 (0.11-0.83) Nor-LSD 3 0.01 (0.01-0.02) 4.6 (4.3-5.0) 11 (8.5-11) 0.21 (0.20 - 0.30) 0.28 (0.28-0.35)

* HO-LSD and HO-LSD glucuronide were qualitatively determined and relative concentrations are given related to peak areas of LSD-d3 of 0.1 ng/mL. The 793 presented values (Cmax and AUC) therefore do not reflect actual absolute concentrations of these analytes. Drug Testing and Analysis A. E. Steuer et al.

commercially available as reference material. Nevertheless, addi- spectra of the suspected HO glucuronide (precursor m/z 516) tional targets that might allow for elongated detection times of (Figure ) matched the postulated HO-LSD spectrum (Figure 2B), LSD consumption are of major interest in clinical and forensic toxi- as was expected due to the initial loss of glucuronic acid. Minor cology. Therefore, plasma samples at expected tmax concentrations peaks for other potential metabolites could be detected, but not re- (t =2.5 h) were initially screened for the presence of other described liably identified by mass spectral interpretation. In former studies, metabolites by an EPI scan experiment and manual spectra inter- these metabolites were only identified by neutral loss experiments pretation. From the seven potential metabolites included in the and predefined MRM traces, but not by unambiguous mass spectral study, only an HO-LSD metabolite and its glucuronide could be un- data.[11] Therefore, besides oxo-HO-LSD and nor-LSD that could be ambiguously identified. The corresponding EPI spectra and postu- validatedly quantified, MRM transitions for HO-LSD and HO-LSD lated fragmentations are given in Figure 2. Compared to LSD, in glucuronide were included into an additional method using the ex- the mass spectrum resulting from the precursor ion of m/z 340 all act same settings as for the validated one. All PK samples were mea- major fragments were shifted by 16 u indicating a sured twice – using the validated quantification method and a on the tetracyclic ring system (Figure 2B). Two HO-LSD isomers hy- second (qualitative/semi-quantitative) method. A chromatogram droxylated in positions 13 and 14 on the indole moiety were pro- of an authentic plasma sample showing presence and baseline sep- posed from initial metabolism studies in rats.[43,44] In humans, aration of HO-LSD glucuronide, oxo-HO-LSD, HO-LSD, nor-LSD, LSD, presence of either 13- or 14-HO-LSD as definite chemical structure and iso-LSD is given in Figure 3. has never been proven. Previous studies in human urine could only identify one peak for HO-LSD without conclusive structure elucida- Pharmacokinetic analysis of LSD, iso-LSD, and metabolites tion 11. Based only on MS analysis, hydroxy isomers on the indole moiety cannot be differentiated. Additionally, hydroxylation of the Limited data on LSD pharmacokinetics are available.[16,25,45] Dolder N-ethyl function to LEO has been described[11] which should be et al. collected blood samples for up to 24 h and additionally deter- easily distinguishable through different fragmentation. In our cur- mined oxo-HO-LSD pharmacokinetics.[16] However, LOQs in the rent measurement, only one peak corresponding to a hydroxy me- used analytical method were 0.1 ng/mL for LSD and oxo-HO-LSD tabolite could be observed. There can be several explanations: (1) being not sensitive enough for quantification of all blood samples only one HO-metabolite is formed in humans, (2) other isomers over the 24 h time period. LSD could only be measured in all 16 par- are of too low abundance for reliable detection, or (3) the applied ticipants up to 12 h post administration, while oxo-HO-LSD chromatography was unable to resolve different HO isomers. Mass concentration-time profiles in general could only be detected in

794 Figure 2. Enhanced product ion (EPI) mass spectra of LSD (A), HO-LSD (B), and HO-LSD glucuronide (C), their corresponding and postulated fragmentation. EPI spectra were recorded applying a collision energy spread (CES) of 35 +/- 15 eV. Drug Testing New microflow LC-MS/MS method for quantification of LSD and its metabolites and application and Analysis

Figure 3. Microflow-MS/MS extracted ion chromatograms of an authentic plasma sample 2 h after LSD administration. Concentrations determined were as follows: LSD 3.8 ng/mL, iso-LSD 1.0 ng/mL, oxo-HO-LSD 0.05 ng/mL and nor-LSD < limit of quantification (LOQ). one-half of the subjects.[16] Analysis with the newly developed also be formed in vivo. Therefore, it remains a point of discussion method allowed for quantification of LSD and oxo-HO-LSD in all whether the presented PK data for iso-LSD are only related to an in- 16 participants at all time points except for participant 2 with both gestion of iso-LSD or to an additional in vivo formation. Nor-LSD analytes detectable > LOQ (0.01 ng/mL) up to 16 h post-dose. Addi- could be detected and quantified in only one subject in plasma tionally, iso-LSD and nor-LSD were validatedly quantified and HO- samples from 1 to 24 h and 2 others from 1 to 12 h post LSD and HO-LSD glucuronide could be identified and concentra- ingestion. In all other participants its detection seemed rather tions could be compared relatively. Iso-LSD concentrations random and/or was below the LOQ. HO-LSD was qualitatively were > LOQ in all samples. Iso-LSD is typically detected as a by- detectable in 11 and HO-LSD glucuronide in 8 subjects till 24 h, product during the LSD synthesis[10] or formed for example under respectively, except for 2 and 1 participant(s) each, where identifi- alkaline conditions.[9] Purity of LSD administered in the current cation was possible up to 16 h. Peak areas of HO-LSD glucuronide study was estimated to be >98%; however, slight formation in generally were much lower than those of HO-LSD. However, it must the capsules over time cannot be completely excluded. Previously be considered that an SPE was conducted prior to analysis, and performed content analyses of identical capsules kept at these con- recovery of glucuronides remains unknown. Calculated PK data ditions showed a decrease in LSD content <3% within 12 months. using non-compartmental analysis for all analytes are given in Iso-LSD contents of approximately 5% could be detected in 24- Table 2 and concentration-time profiles are depicted in Figure 4. month-old capsules. Metabolites of iso-LSD similar to LSD were In general, PK data were comparable to previous results.[16] Interest- not detected but might have been below the detection limits of ingly, LSD and iso-LSD showed different PK with longer elimination the method. The method validation process provided no indication half-life for iso-LSD (median 12 h) compared to LSD (median 4.2 h). for formation of iso-LSD from LSD during the extraction procedure In general, elimination of metabolites was slower than for LSD. Due or processed sample storage. Long-term stability (6 months) has to very low concentrations of all LSD metabolites in plasma, been shown for LSD when kept under frozen conditions.[14,23] So prolongation of detection windows by monitoring of these metab- far, in vivo formation of iso-LSD as an LSD metabolite has not been olites seems unlikely, especially taking into account that a sensitive discussed. The presence of a small amount of iso-LSD in the cap- and specifically developed method was necessary to detect those sules in the current study hinders the hypothesis that iso-LSD can over 24 h.

Figure 4. Concentration-time profiles of LSD, iso-LSD and nor-LSD (left side) and oxo-HO-LSD, HO-LSD and HO-LSD glucuronide (right side). HO-LSD and HO- 795 LSD glucuronide were qualitatively determined and relative concentrations are given related to peak areas of LSD-d3 of 0.1 ng/mL. Drug Testing and Analysis A. E. Steuer et al.

Conclusions [14] K. L. Klette, C. K. Horn, P. R. Stout, C. J. Anderson. LC-MS analysis of human urine specimens for 2-oxo-3-hydroxy LSD: Method validation for potential interferants, and stability study of 2-oxo-3-hydroxy LSD We present an MFLC-MS/MS method for the validated quantifica- under various storage conditions. J. Anal. Toxicol. 2002, 26, 193. tion of LSD, iso-LSD, oxo-HO-LSD, and nor-LSD in plasma. MFLC [15] Z. Y. Li, A. J. McNally, H. Y. Wang, S. J. Salamone. Stability study of LSD allowed for sensitive detection and fast analysis. The method ful- under various storage conditions. J. Anal. Toxicol. 1998, 22, 520. filled the required validation criteria and was successfully applied [16] P. C. Dolder, Y. Schmid, M. Haschke, K. M. Rentsch, M. E. Liechti. to the analysis of plasma samples from a controlled LSD administra- Pharmacokinetics and concentration-effect relationship of oral LSD in humans. Int. J. Neuropsychopharmacol. 2016, 19, pyv072. tion study to humans. Determined LSD pharmacokinetic results [17] A. Chung, J. Hudson, G. McKay. Validated ultra-performance liquid were comparable to previous measurements. Oxo-HO-LSD kinetics chromatography-tandem mass spectrometry method for analyzing could be determined in all but one participant over 24 h. Nor-LSD LSD, iso-LSD, nor-LSD, and O-H-LSD in blood and urine. J. Anal. and additionally HO-LSD and HO-LSD glucuronide were detected Toxicol. 2009, 33, 253. for the first time in human plasma. Although half-lives of all metab- [18] H. Schutz, A. Paine, F. Erdmann, G. Weiler, M. A. Verhoff. Immunoassays for drug screening in urine : Chances, challenges, and pitfalls. Forensic olites exceeded those of LSD, screening for metabolites to increase Sci. Med. Pathol. 2006, 2, 75. detection windows in plasma seems not to be constructive due to [19] E. D. Clarkson, D. Lesser, B. D. Paul. Effective GC-MS procedure for their very low concentrations. detecting iso-LSD in urine after base-catalyzed conversion to LSD. Clin. Chem. 1998, 44, 287. [20] D. Libong, S. Bouchonnet, I. Ricordel. A selective and sensitive method for quantitation of lysergic acid diethylamide (LSD) in whole blood by Acknowledgements gas chromatography-ion trap tandem mass spectrometry. J. Anal. Toxicol. 2003, 27, 24. The authors would like to thank their colleague Sandra Staeheli for [21] P. C. Dolder, M. E. Liechti, K. M. Rentsch. Development and validation of her support and discussion and express their gratitude to Emma a rapid turboflow LC-MS/MS method for the quantification of LSD and Louise Kessler, MD, for the generous legacy she donated to the In- 2-oxo-3-hydroxy LSD in serum and urine samples of emergency stitute of Forensic Medicine at the University of Zurich, Switzerland toxicological cases. Anal. Bioanal. Chem. 2015, 407, 1577. for research purposes. [22] D. Favretto, G. Frison, S. Maietti, S. D. Ferrara. LC-ESI-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. 2007, 121, 259. References [23] R. Martin, J. Schurenkamp, A. Gasse, H. Pfeiffer, H. Kohler. Determination of , bufotenine, LSD and its metabolites in [1] T. Passie, J. H. Halpern, D. O. Stichtenoth, H. M. Emrich, A. Hintzen. The serum, plasma and urine by SPE-LC-MS/MS. Int. J. Legal Med. 2013, pharmacology of lysergic acid diethylamide: a review. CNS Neurosci. 127, 593. Ther. 2008, 14, 295. [24] J. H. Sklerov, J. Magluilo Jr., K. K. Shannon, M. L. Smith. Liquid [2] D. E. Nichols. Psychedelics. Pharmacol. Rev. 2016, 68, 264. chromatography-electrospray ionization mass spectrometry for the [3] T. S. Krebs, P. O. Johansen. Over 30 million psychedelic users in the detection of lysergide and a major metabolite, 2-oxo-3-hydroxy-LSD, United States. F1000Res. 2013, 2, 98. in urine and blood. J. Anal. Toxicol. 2000, 24, 543. [4] Y. Schmid, F. Enzler, P. Gasser, E. Grouzmann, K. H. Preller, [25] G. K. Aghajanian, O. H. Bing. Persistence of lysergic acid diethylamide in F. X. Vollenweider, R. Brenneisen, F. Muller, S. Borgwardt, M. E. Liechti. the plasma of human subjects. Clin. Pharmacol. Ther. 1964, 5, 611. Acute effects of lysergic acid diethylamide in healthy subjects. Biol. [26] X. He, M. Kozak, S. Nimkar. Ultra-sensitive measurements of 11-Nor- Psychiatry. 2015, 78,544. delta(9)--9- in oral fluid by [5] R. L. Carhart-Harris, M. Kaelen, M. Bolstridge, T. M. Williams, microflow liquid chromatography-tandem mass spectrometry using a L. T. Williams, R. Underwood, A. Feilding, D. J. Nutt. The paradoxical benchtop quadrupole/orbitrap mass spectrometer. Anal. Chem. 2012, psychological effects of lysergic acid diethylamide (LSD). Psychol. 84, 7643. Med. 2016, 46, 1379. [27] M. Concheiro, D. Lee, E. Lendoiro, M. A. Huestis. Simultaneous [6] P. C. Dolder, Y. Schmid, F. Muller, S. Borgwardt, M.E. Liechti. LSD acutely quantification of Delta(9)-tetrahydrocannabinol, 11-nor-9-carboxy- impairs fear recognition and enhances emotional empathy and tetrahydrocannabinol, cannabidiol and in oral fluid by sociality. Neuropsychopharmacology 2016.doi:10.1038/npp. microflow-liquid chromatography-high resolution mass spectrometry. [7] M. Klein, F. Kramer. Rave drugs: pharmacological considerations. AANA J. Chromatogr. A. 2013, 1297, 123. J. 2004, 72, 61. [28] C. C. Christianson, C. J. Johnson, S. R. Needham. The advantages of [8] R. Penning, J. L. Veldstra, A. P. Daamen, B. Olivier, J. C. Verster. Drugs of microflow LC-MS/MS compared with conventional HPLC-MS/MS for abuse, driving and traffic safety. Curr. Drug Abuse Rev. 2010, 3, 23. the analysis of methotrexate from human plasma. Bioanalysis. 2013, [9] S. J. Salamone, Z. Li, A. J. McNally, S. Vitone, R. S. Wu. Epimerization 5, 1387. studies of LSD using 1H nuclear magnetic resonance (NMR) [29] H. Wang, P. Bennett. Performance assessment of microflow LC combined spectroscopy. J. Anal. Toxicol. 1997, 21, 492. with high-resolution MS in bioanalysis. Bioanalysis. 2013, 5, 1249. [10] S. S. Johansen, J. L. Jensen. Liquid chromatography-tandem mass [30] A. E. Steuer, M. Poetzsch, M. Koenig, E. Tingelhoff, S. N. Staeheli, spectrometry determination of LSD, ISO-LSD, and the main A. T. Roemmelt, T. Kraemer. Comparison of conventional liquid metabolite 2-oxo-3-hydroxy-LSD in forensic samples and application chromatography-tandem mass spectrometry versus microflow liquid in a forensic case. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. chromatography-tandem mass spectrometry within the framework of 2005, 825, 21. full method validation for simultaneous quantification of 40 [11] J. Canezin, A. Cailleux, A. Turcant, A. Le Bouil, P. Harry, P. Allain. antidepressants and neuroleptics in whole blood. J. Chromatogr. A. Determination of LSD and its metabolites in human biological fluids 2015, 1381, 87. by high-performance liquid chromatography with electrospray [31] E. Communities. Commision decision of 12 August 2002 implementing tandem mass spectrometry. J. Chromatogr. B Biomed. Sci. App. 2001, Council Directive 96/23/EC concerning the performance of analytical 765, 15. methods and the interpretation of results. Off J Eur Community. 2002, [12] K. L. Klette, C. J. Anderson, G. K. Poch, A. C. Nimrod, M. A. ElSohly. L221,8. Metabolism of lysergic acid diethylamide (LSD) to 2-oxo-3-hydroxy [32] F. T. Peters, M. Hartung, M. Herbold, G. Schmitt, T. Daldrup, F. Musshoff. LSD (O-H-LSD) in human microsomes and cryopreserved human Appendix B to the GTFCh guidelines for quality assurance in forensic- hepatocytes. J. Anal. Toxicol. 2000, 24, 550. toxicological analyses. Requirements for the validation of analytical [13] G. Skopp, L. Potsch, R. Mattern, R. Aderjan. Short-term stability of methods. Toxichem. Krimtech. 2009, 76, 185. lysergic acid diethylamide (LSD), N-desmethyl-LSD, and 2-oxo-3- [33] B. K. Matuszewski, M. L. Constanzer, C. M. Chavez-Eng. Strategies for the 796 hydroxy-LSD in urine, assessed by liquid chromatography-tandem assessment of matrix effect in quantitative bioanalytical methods mass spectrometry. Clin. Chem. 2002, 48, 1615. based on HPLC-MS/MS. Anal. Chem. 2003, 75, 3019. Drug Testing New microflow LC-MS/MS method for quantification of LSD and its metabolites and application and Analysis

[34] H. H. Maurer, K. Pfleger, A. A. Weber. Mass Spectral and GC Data of [40] R. C. Baselt. Disposition of Toxic Drugs and Chemicals in Man, Drugs, Poisons, Pesticides, Pollutants and their Metabolites, Wiley-VCH, 8thEdnChemical Toxicology Institute, Foster City CA2008. Weinheim2011. [41] J. Cai, J. Henion. Elucidation of LSD metabolism by liquid [35] V. P. Shah, K. K. Midha, J. W. Findlay, H. M. Hill, J. D. Hulse, I. J. McGilveray, chromatography and capillary electrophoresis coupled with tandem G. McKay, K. J. Miller, R. N. Patnaik, M. L. Powell, A. Tonelli, mass spectrometry. J. Anal. Toxicol. 1996, 20, 27. C. T. Viswanathan, A. Yacobi. Bioanalytical method validation--a revisit [42] G. K. Poch, K. L. Klette, D. A. Hallare, M. G. Manglicmot, R. J. Czarny, with a decade of progress. Pharm. Res. 2000, 17, 1551. L. K. McWhorter, C. J. Anderson. Detection of metabolites of lysergic [36] F. T. Peters, O. H. Drummer, F. Musshoff. Validation of new methods. acid diethylamide (LSD) in human urine specimens: 2-oxo-3-hydroxy- Forensic Sci. Int. 2007, 165, 216. LSD, a prevalent metabolite of LSD. J. Chromatogr. B Biomed. Sci. Appl. [37] F. T. Peters, H. H. Maurer. Bioanalytical method validation and its 1999, 724,23. implications for forensic and clinical toxicology - A review. Accredit. [43]Z.H.Siddik,R.D.Barnes,L.G.Dring,R.L.Smith,R.T.Williams. Qual. Assur. 2002, 7, 441. The metabolism of lysergic acid DI[14C]ethylamide ([14C]LSD) [38] A. Kruve, R. Rebane, K. Kipper, M. L. Oldekop, H. Evard, K. Herodes, in the isolated perfused rat liver. Biochem. Pharmacol. 1979, P. Ravio, I. Leito. Tutorial review on validation of liquid 28, 3081. chromatography-mass spectrometry methods: part I. Anal. Chim. [44] Z. H. Siddik, R. D. Barnes, L. G. Dring, R. L. Smith, R. T. Williams. The fate Acta. 2015, 870, 29. of lysergic acid DI[14C]ethylamide ([14C]LSD) in the rat, guinea pig and [39] A. Kruve, R. Rebane, K. Kipper, M. L. Oldekop, H. Evard, K. Herodes, rhesus monkey and of [14C]iso-LSD in rat. Biochem. Pharmacol. 1979, P. Ravio, I. Leito. Tutorial review on validation of liquid 28, 3093. chromatography-mass spectrometry methods: part II. Anal. Chim. [45] D. G. Upshall, D. G. Wailling. The determination of LSD in human Acta. 2015, 870,8. plasma following . Clin. Chim. Acta. 1972, 36, 67. 797