An LC-MSn method was developed for selective detection of 37 tryptamines, five β-carbolines, ibogaine, and yohimbine in human urine and plasma, and for validated quantification of 33 of them in plasma. Eight tryptamines and the five β-carbolines did not fulfill validation criteria but could at least be identified.

2 and MS3 and matching with the reference spectra of the library [ 40 2 Abbreviated drug name Contained in solution no. (compound no. in Fig. 2) LOD urine [ng/mL] after PP and LC-MSn LOD plasma [ng/mL] after LLE and LC-MSn 4-HO-MET 1 (1) 100 100 NMT 1 (2) 100 100 Harmol 1 (3) 100 100 5-MeO-2-Me-DMT 1 (4) 100 100 Harmaline 1 (5) 10 1 4-AcO-DiPT 1 10 100 5-MeO-2-Me-DiPT 1 (7) 10 1 5-MeO-2-Me-2-MALET 1 (8) 10 1 5-BnO-DMT 1 (9) 10 100 2-Ph-DiPT 1 (10) 100 100 5-BnO-DPT 1 (11) 10 100 Harmalol 2 (12) 10 100 4-HO-DiPT 2 (13) 100 100 MiPT 2 (14) 100 100 5-MeO-2-Me-MiPT 2 (15) 100 100 5-MeO-2-Me-EiPT 2 (16) 10 1 5,6-MD-DALT 2 (17) 100 100 DALT 2 (18) 10 100 Ibogaine 2 (19) 10 1 DPT 2 (20) 100 100 5-EtO-DALT 2 (21) 100 100 5-MeO-2-Me-ALCHT 2 (22) 10 1 5-BnO-DALT 2 (23) 10 100 4-HO-MiPT 3 (24) 100 100 NiPT 3 (25) 100 100 5-MeO-2-Me-Pyr-T 3 (26) 10 100 Harmine 3 (27) 10 1 Yohimbine 3 (28) 10 1 5-MeO-DPT 3 (29) 100 100 2-Ph-DMT 3 (30) 100 100 5-MeO-2-Me-DPT 3 (31) 100 100 5-Me-DALT 3 (32) 100 100 7-Et-DALT 3 (33) 100 100 2-Ph-DALT 3 (34) 10 100 Bufotenin 4 (35) 100 100 Psilocin 4 (36) 100 100 DMT 4 (37) 100 100 Tetrahydroharmine 4 (39) 100 100 5-MeO-2-Me-Pip-T 4 (40) 100 100 5-MeO-DALT 4 (41) 100 100 5-MeO-2-Me-EPT 4 (42) 10 100 5-MeO-2-Me-DALT 4 (43) 100 1 7-Me-DALT 4 (44) 100 100 5-BnO-DiPT 4 (45) 10 100 As shown in previous studies, the standard urine screening approach (SUSA) allowed the detection of all drugs in human urine, if not completely metabolized, based on full mass spectra recorded in the stages of MSand MSand matching with the reference spectra of the library []. Detection in plasma was based on the recorded full PIS spectra of previously defined target compounds (protonated molecules) and subsequent library match based on the same settings used for urine. The identification via library-based identification allowed automated data file evaluation with reliable detection of the tested compounds and distinguishing the structurally similar isomers, which would not be possible solely using transitions. The method was selective using both strategies, since eight blank urine and plasma samples from different sources did not show any interfering peaks influencing the analytes or the IS signals. In addition, zero samples showed no interferences as well. The resulting LODs for urine and plasma are summarized in Table

After detection in urine or plasma samples, quantification in human plasma could additionally be performed with LLOQs set to 1.5 ng/mL (QC low) for each analyte. These limits seemed to be reasonable since plasma levels reported after ingestion of such compounds were between 6.2 and 115 ng/mL [27]. The calibration for all analytes was calculated finally using a linear fitting with a 1/x 2 weighting after visual inspection of individual residual plots. Using this well-established standard approaches, applicability of the approach could be maintained in emergency toxicology as well as routine work.

RE, ME, and PE experiments with six different plasma sources at low and high concentrations were performed, and RE, ME, and PE were calculated by area comparison as proposed by Matuszewski et al. [48]. The liquid–liquid extraction (LLE) method used was previously established and shown to be applicable for quantification of a wide range of compounds [38]. As there were no definite AC for the allowed variations of RE experiments, the acceptance limit was set to 20 % CV for low and 15 % CV for high concentrations. In the authors’ opinion, the CV was deemed most suitable to characterize variation of RE values between different plasma samples and different analytes. Due to the nature of sample preparation (transfer of 500 μL supernatant of the used butyl acetate/ethyl acetate mixture), the maximum, theoretically obtainable RE was 83.3 %. The AC for ME has been set to 75–125 % according to previous reports [50, 51, 52]. A CV of 15 % and 20 % (near the limit of detection) was considered acceptable. As PE is a combination of RE and ME tests, the acceptance limit for the CV was based on RE.

The median RE was 74 (29–102) % for QC low and 68 (20–99) % for QC high, whereas the CV of 12 of the 44 analytes (QC low) and 12 of the 44 analytes (QC high) were above the AC. 5-BnO-DPT, harmalol, 5-BnO-DiPT, 5-MeO-DALT, 5-MeO-2-Me-DALT, 5-MeO-2-Me-EPT, 7-Me-DALT, bufotenin, and psilocin were above the AC for both concentrations. For 4-AcO-DiPT, harmalol, harmine, bufotenin, and psilocin, RE of QC low could not be determined due to insufficient sensitivity resulting in an insufficient number of data points across the peaks. In summary, and considering the fact that the maximum RE that could be obtained was 83.3 %, the all over all REs gained for this LLE were deemed acceptable.

53 37 2 3 Open image in new window Abbreviated name Used IS QC Low (nominal concentration 1.5 ng/mL) QC Med (nominal concentration 50 ng/mL) QC High (nominal concentration 95 ng/mL) CM Bias REP CV IMP CV CM Bias REP CV IMP CV CM Bias REP CV IMP CV NMT Zolpidem-d 6 1.5 1.4 16 16 50 1.2 7.6 9.2 89 −5.8 6.1 9.5 DMT − − − − − − − − − − − − − Harmol Zolpidem-d 6 − − − − 48 −2.8 12 13 86 −9.5 11 11 Harmalola Zolpidem-d 6 − − − − 50 1.0 78 68 88 −6.5 50 53 NiPT Citalopram-d 6 1.5 −0.2 15 20 49 −2.7 5.1 10 82 −13 8.0 10 Bufotenina − − − − − − − − − − − − − Psilocina − − − − − − − − − − − − − Harminea Citalopram-d 6 − − − − 47 −5.1 29 30 93 −1.5 13 24 Harmaline Zolpidem-d 6 − − − − 41 −16 8.8 26 81 −14 12 14 MiPT Zolpidem-d 6 1.4 −8.2 10 18 48 −3.4 10 10 85 −9.7 9.2 13 Tetrahydroharmine Zolpidem-d 6 1.4 −5.3 27 29 46 −6.7 6.1 8.7 85 −10 8.1 10 4-HO-MET Citalopram-d 6 1.2 −19 16 17 46 −7.0 13 12 84 −10 7.1 10 4-HO-MiPT Citalopram-d 6 1.4 −8.5 16 18 47 −4.3 8.4 12 81 −14 9.8 11 5-MeO-2-Me-DMT Zolpidem-d 6 1.5 −2.8 10 16 48 −3.2 8.4 10 92 −2.5 4.3 9.6 DALT Citalopram-d 6 1.3 −11 9.7 11 49 −1.7 5.9 8.6 86 −9.3 7.9 8.6 DPT Citalopram-d 6 1.4 −8.4 8.5 8.6 47 −4.8 6.3 9.3 81 −14 4.3 6.9 5-Me-DALT Citalopram-d 6 1.3 −13 17 19 48 −2.5 5.3 9.9 82 −13 7.9 7.4 7-Me-DALTa Citalopram-d 6 1.6 3.5 11 17 49 −0.8 5.3 6.8 87 −7.6 5.1 7.4 5-MeO-2-Me-Pyr-T Citalopram-d 6 1.4 −6.3 13 16 50 0.0 7.5 12 81 −14 7.0 8.4 4-HO-DiPT Zolpidem-d 6 1.2 −19 18 19 48 −3.3 11 13 87 −7.9 7.2 13 5-MeO-2-Me-MiPT Zolpidem-d 6 1.4 −7.8 11 15 48 −2.2 11 14 84 −11 11 14 2-Ph-DMT Citalopram-d 6 1.3 −11 17 16 50 0.5 6.4 7.4 93 −1.6 5.7 7.7 7-Et-DALT Citalopram-d 6 1.5 −2.3 17 18 50 0.4 4.5 10 81 −14 8.0 9.2 5-MeO-DALTa Citalopram-d 6 1.3 −10 10 19 47 −5.2 6.3 8.3 89 −5.3 4.3 9.0 5-MeO-2-Me-Pip-T Zolpidem-d 6 1.7 10 9.4 14 49 −0.4 7.9 11 93 −1.5 11 11 5-MeO-DPT DiPT-d 4 1.6 5.9 19 18 50 0.1 3.3 4.3 84 −11 9.1 7.3 5-MeO-2-Me-EPTa Citalopram-d 6 1.4 −8.2 13 18 47 −5.7 9.1 10 81 −14 8.2 9.8 5-MeO-2-Me-EiPT Zolpidem-d 6 1.4 −6.4 14 19 51 3.0 13 13 89 −6.3 9.5 13 5,6-MD-DALT Citalopram-d 6 1.2 −20 11 12 46 −7.3 8.6 12 85 −10 10 11 5-MeO-2-Me-DALTa Citalopram-d 6 1.4 −4.1 9.5 15 48 −2.5 7.2 8.3 87 −7.9 7.2 11 5-EtO-DALT Citalopram-d 6 1.3 −15 9.5 11 48 −2.9 5.9 9.6 87 −8.1 6.8 7.8 5-MeO-2-Me-2-MALET Zolpidem-d 6 1.5 0.0 12 15 50 1.1 6.3 9.2 95 0.4 7.2 12 5-MeO-2-Me-DPT Citalopram-d 6 1.5 −0.9 15 19 49 −1.2 6.4 10 82 −12 9.4 8.1 5-MeO-2-Me-DiPT Zolpidem-d 6 1.5 2.0 11 16 51 2.0 11 12 97 2.2 7.4 12 5-BnO-DMT Zolpidem-d 6 1.7 16 17 17 48 −2.7 5.0 7.7 94 −0.9 10 14 4-AcO-DiPTa − − − − − − − − − − − − − Ibogaine Citalopram-d 6 1.3 −14 11 20 47 −4.2 5.1 10 84 −10 8.3 9.3 2-Ph-DALT Citalopram-d 6 1.4 −5.4 19 28 47 −5.5 10 11 79 −15 11 15 2-Ph-DiPT Citalopram-d 6 1.6 4.4 15 16 50 1.7 9.2 11 97 2.1 11 15 5-MeO-2-Me-ALCHT 5-EtO-ALCHT-d 4 1.4 −6.7 10 13 50 1.2 8.7 10 87 −8.1 12 15 5-BnO-DALT 5-EtO-ALCHT-d 4 1.4 −5.4 13 31 55 10 11 28 94 −0.8 9.6 24 5-BnO-DPTa Trimipramin-d 3 1.3 −12 20 18 50 1.0 15 12 90 −4.9 10 15 5-BnO-DiPTa Trimipramin-d 3 2.4 56 29 29 56 13 49 50 90 −5.2 25 28 Yohimbine Citalopram-d 6 1.4 −4.3 17 20 48 −2.3 7.7 10 80 −15 7.3 10 As ion suppression and enhancement effects of co-eluting analytes in multi-analyte approaches are a well-documented phenomenon [] and also observed in the presented study, the 44 analytes were separated equally into four solutions to overcome this issue. Each solution was prepared and analyzed as described above. Furthermore, compounds often occurring as drugs of abuse in patient samples such as amphetamine and cocaine as well as some common over-the-counter drugs were also tested for ion suppression due to co-elution. Fortunately, they were either separated chromatographically or did not show any influence on the analytes. In order to maintain a convenient instrumental setup usable for routine applications, the LC gradient employed for the separation of analytes in urine and plasma samples was the same as proposed for the SUSA []. Representative chromatograms of all four solutions with analytes at 95 ng/mL are depicted in Fig.. ME with CVs above the AC were observed for nearly all compounds with a median ME of 141 (88–224) % for QC low and 133 (72–343) % for QC high. The CV at QC low ranged from 7 % (yohimbine) to 49 % (5-BnO-DMT) and at QC high from 6 % (bufotenin) to 46 % (DMT). As already mentioned for RE, ME could not be determined for all analytes at QC low due to insufficient sensitivity resulting in a limited number of data points usable for peak integration. Compounds with values for RE and ME above the AC are marked in Table

Results of accuracy, intra-day precision (RSD R , repeatability), and the inter-day precision (RSD T, intermediate precision; between-day precision) were calculated as recommended [46, 47, 54] and are listed in Table 3. With the exception of bufotenin, psilocin, DMT, harmol, harmalol, harmaline, harmine, tetrahydroharmine, 2-Ph-DALT, 5-BnO-DALT, 5-BnO-DiPT, and 4-AcO-DiPT, all analytes (33 out of 44) fulfilled the validation criteria despite the presence of matrix effects, which underlined the necessity for matrix based calibration. Some of these analytes were already above AC during RE and ME experiments. On the other hand, 4-HO-MET, 5-BnO-DPT, 5-MeO-2-Me-EiPT, ibogaine, and MiPT, not fulfilling the set criteria during RE experiments, showed acceptable precision and accuracy results. Further reasons might be stability problems for bufotenin and psilocin during validation, which were unstable in processed samples and even stock solutions. Furthermore, the MSn method did not provide sufficient sensitivity and hence sufficient amount of scan points for thorough quantification of the five β-carbolines (harmol, harmalol, harmaline, harmine, and tetrahydroharmine) and DMT leading to insufficient results in the range of QC low. For 5-BnO-DALT and 2-Ph-DALT, the inter-day precision of QC low, CQ med, and QC high did not fulfill the validation criteria. For 5-BnO-DiPT, the inter-day and intra-day precision as well as bias of QC low were not sufficient to match the criteria. Quantification of 4-AcO-DiPT was not possible due to in-source deacylation to 4-HO-DiPT.

Accuracy and precision data for LLOQ were within in the AC of 20 % [49], with exception of 4-AcO-DiPT, harmaline, harmol, 5-BnO-DALT, 5,6-MD-DALT, harmalol, 2-Ph-DALT, harmine, 5-BnO-DiPT, bufotenin, DMT, psilocin, and tetrahydroharmine. As the AC of 20 % accuracy and precision data at LLOQ for the above-mentioned analytes were not fulfilled at any concentration levels, no LLOQ could be determined. For these analytes, only qualitative statements could be made. Stability studies revealed that all analytes were stable in extracts for 48 h at +8 °C, except for psilocin and bufotenin. Analytes, with exception of psilocin and bufotenin, were also stable in methanolic stock solutions for at least 4 weeks. Additionally, stability of analytes could be prolonged using DMSO as solvent and ascorbic acid as preservative [28].