Development of a liquid chromatography - high resolution mass spectrometry method for multi-component screening of synthetic cannabinoids in blood.

Örebro University

School of Health and Medical Sciences

Department of Clinical Medicine

Program: Biomedicine and Methods in Medical Diagnostic: Experimental Medicine, 120 ects

Course: Degree Project in Medicine, 45 etcs

Date: May 18

th

₂₀₁₆

Development of a liquid chromatography - high resolution mass

spectrometry method for multi-component screening of synthetic

cannabinoids in blood.

Author: Luiza Tworek

Supervisor: Olof Beck, professor

Division of

Clinical Pharmacology,

Karolinska University Hospital

Anders Helander, professor

Division of Clinical Chemistry,

Karolinska University Hospital

ABSTRACTDetection of NPS remains an analytical challenge, due to a great variety of compounds and rapid exchange of the products available on the recreational drugs market. In recent years, several methods for detection of synthetic cannabinoids based on liquid chromatography (LC) combined with mass spectrometry (MS) have been developed. Most of these methods utilize com-plicated sample preparation methods and a targeted approach for each individual compound. We developed a screening method for the simultaneous detection of 43 synthetic cannabinoids in serum, by LC and high-resolution mass spectrometry (HRMS). We used a simplified sample cleaning procedure, protein precipitation, to minimize the cost of the analysis and make it more environment-friendly. Data were acquired in a full scan mode. Analytes were identified and quantified with reference standards spiked in drug-free plasma, that were separated on an Accucore C18 column using a Dionex UltiMate 3000 UHPLC system,Thermo Fisher Scien-tific and detected by an OrbiTrap Q Exactive, Thermo Fisher ScienScien-tific. 38 analytes fulfilled the requirement for quantitative meth-od and were validated according to an internal protocol. The experiments included determination of the limit of detection, limit of quantification, linearity, matrix effect, recovery, precision and accuracy. Remaing 5 analytes were still included in the method, in a qualitative maner. The method was applied to 84 authentic case samples, that resulted in identification of 16 different synthetic cannabinoids. The method is rapid, sensitive and highly specific for variety of synthetic cannabinoids.

The use of Cannabis sativa, a herbal plant used for preparation of marijuana, has been a point of interest for centuries. Since the structural formula of ∆-tetrahydroxycannabinol (THC), a psychoactive substance of the cannabis plant, was determined and its mechanism of action described, synthetic cannabinoid analogs have been synthesized in order to evaluate their poten-tial as therapeutic agents. [1]

The pharmacological effects of THC and synthetic canna-binoids are due to their action on cannabinoid receptors (CB) 1 and 2. CB1 occurs in the brain, neuronal cells and tissues, and is mainly presented on central and peripheral nerve terminals. This localization corresponds with the behavioral effects of CB1 agonists, such as anxiety and panic. The CB2 receptor is predominantly presented in the immune system, in the spleen, lymph nodes and tonsils and has been suggested to play a role in pain and control of emesis. [2-6]CB2 agonists have been considered as possible targets in inflammation pain.

JWH-018 (1-pentyl-3-(1-naphtoyl) indole), created by J. W. Huffman at Clemenson University, was one of the first synthe-sized CB2 agonist to investigate those possibilities. Huffman and colleagues showed that the affinity of JWH-018 was 10-fold higher compared to THC. [7] Unfortunately, synthetic cannabinoids turned out to be more toxic than beneficial. Because of their high potency, unknown drug-drug interaction, the effect of their action is hard to predict.

In 2004, synthetic cannabinoids were introduced as a legal alternative to cannabis, under the brand name “Spice”. Spice was a herbal mixture intended for smoking, that was obtained in head shops or through internet sale. [8] The content and quality of Spice mixtures, together with unknown pharmaco-dynamics and toxicological effects, made the clinical effects unpredictable. Users of Spice have presented with a range of life threatening adverse effects such as hypertension, visual and auditory hallucinations, psychosis, intracranial hemor-rhage and even death. [2, 5]

Since the appearance of Spice on the internet, its popularity and variety has increased dramatically. [8] The number of confiscated “designer drugs” in Europa, nowadays commonly referred to as “new psychoactive substances” (NPS) not only including synthetic cannabinoids, increased from 126 to 450 from 2009 to 2014. Only in 2014, 177 different synthetic cannabinoids (39% of all reported NPS) were reported to the United Nations Office on Drugs and Crime (UNODC) early warning advisory. [9] According to reports from the Swedish Council for Information on Alcohol and Other Drugs (CAN) 2014 and 2015 reports, synthetic cannabinoids are the third largest narcotic group (after two cannabis products) used by students in Swedish high schools. [10, 11]

Detection of NPS remains an analytical challenge, due to a great variety of compounds and rapid exchange of the prod-ucts available on the recreational drugs market. Manufacturers immediately release a new variant, when one substance is controlled. In recent years, several methods for detection of synthetic cannabinoids based on liquid chromatography (LC) combined with mass spectrometry (MS), including LC with single (LC-MS), tandem (LC-MS/MS), or high-resolution MS (LC-HRMS), have been developed. [12-19] Most of these methods utilize liquid-liquid extraction and a targeted ap-proach for each individual compound. Those strategies rely on reference standards, and multiple reaction monitoring or li-brary search, and are therefore not useful to detect newly designed synthetic analogs. HRMS gives this possibility as

there is no necessity for reference substances or knowledge of target masses and structures. New compounds can be intro-duced any time, and using full-scan mode allows for retrospec-tive evaluation of previously collected data. A disadvantage is that the HRMS instruments require more sophisticated data-analysis software.[3, 20]

Our group decided to use a compromise between an old-fashion targeted approach and the novel HRMS technique, to develop an MS-based screening method for 43 synthetic can-nabinoids. The goal was to use a simplified sample cleaning procedure to lower the cost of analysis, and make it more rapid and environment-friendly.

MATERIAL AND METHODS

Reagents and Standards. All solvents were of analytical or

HPLC grade. Methanol and acetonitrile were purchased from Fischer, formic acid (98-100%) was from Merck, and ammo-nium formate of LC-MS Ultra grade was obtained from Fluka. The water used for the experiments was of MilliQ Ultra Quali-ty (>18 MΩ/cm).

Most reference materials of synthetic cannabinoids were pur-chased from Cayman (USA), whereas some were kindly pro-vided by the National Forensic Centre (NFC; Linköping, Swe-den). The reference materials obtained from non-commercial suppliers has a certificate that confirm authentic substance content and purity grade.

Preparation of Stock Solutions. Stock solutions for all

refer-ence standards were prepared by dissolving a weighted amount in methanol. All liquid reference standards, including internal standard stock solutions, were used directly for the preparation of working solutions.

Preparation of Working Solutions. The working solutions

were prepared in methanol in five mixes at concentrations of 20 µg/mL of each substance.

Mix 1: AM1220, JWH015, CB13, URB597, AB001, AM2233.

Mix 2: AKB48, A834735, MAM2201, JWH147, JWH203, JWH307, AM2201.

Mix 3: JWH398, JWH210, STS135, BB22, RSC2, 5F-PB22, PB22, 5F-AKB48.

Mix 4: MN18, EG018, JWH098, 5F-AMB, JWH020, XLR11, AM694, JWH251.

Mix 5: THJ 018, MDMB-CHMICA, ADB-FUBINACA, FUBINACA, EMB-FUBINACA, PINACA, AB-CHMINACA, 5F-PY-PINACA, UR144, 5F-ADB-PINACA, MAB-CHMINACA, FUB-AMB, MMB2201, ADBICA. A final working solution, at 200 ng/mL in methanol and stored at -80°C in 300-µL portions, was prepared by diluting 50 µL of mix 1–5 to a final volume of 5 mL.

A 200-ng/mL internal standard solution (IS) was prepared in 10 mmol/L ammonium format containing 0.005% formic acid.

Preparation of Solutions and Mobile Phases. The

precipita-tion soluprecipita-tion was prepared by mixing methanol and acetoni-trile (50:50, v/v). The solution was stored at room temperature, during the experiment period.

The reconstruction solution was prepared by mixing mobile phase A and mobile phase B (70:30, v/v). The solution was mixed every time new mobile phases were prepared.

Ammonium formate buffer (1 mol/L containing 0.5% formic acid) was prepared by dissolving 6.306 g in 100 mL 0.005% formic acid in water. The buffer was stored at 4°C for a max-imum 14 days.

For preparation of mobile phase A (10 mmol/L ammonium formate with 0.005% formic acid, pH 4.8), 10 mL of ammoni-um formate buffer and 990 mL of MilliQ water (99:1, v/v) were mixed. Mobile phase B (10 mmol ammonium formate with 0.005% formic acid in methanol, pH 4.8) was prepared by mixing 10 mL of ammonium formate buffer, 90 mL of MilliQ water, and 900 mL methanol (90:9:1, v/v/v).

The mobile phases were used and stored at room temperature (RT) for 3 days at most.

Human Plasma and Serum. Five different batches of human

plasma and 2 different batches of serum were used for the experiments (supplied by the Transfusion Medicine Laborato-ry at the Karolinska University Hospital, Huddinge). Both plasma and serum were stored at -20°C in 50-mL portions prior to use.

Clinical samples from 84 patients collected within the Swedish STRIDA project (Samverkansprojekt kring Toxicitetutredning och Riskbedömning av InternetDroger baserat på Laborato-rieanalyser) between years 2010 and 2013 were used for this project. The STRIDA project is conducted in accordance with the Helsinki Declaration and is approved by the regional ethi-cal review board (No. 2013/116–31/2).

Sample Preparation. For analysis, 100 µL of plasma or

se-rum was used following protein precipitation. The sample was fortified with 10 µL of IS and added with 400 µL of precipita-tion soluprecipita-tion. The mixture was vortexed for 30 s and centri-fuged at 4350 RPM for 15 min (Multifuge 3s Heraeus, Ger-many). Of the supernatant, 380 µL were transferred to a new tube and evaporated to dryness using a vacuum centrifuge at 50°C for 1.5 h or under a gentle stream of air at 40°C. The dried sample was reconstructed in 50 µL reconstruction solu-tion and vortexed for 30 s. The reconstructed sample was finally transferred to a new vial and centrifuged another time at 4350 RPM for 5 min.

LC–Electrospray Ionization (ESI) HRMS. Experiments

were performed using an LC-ESI-HRMS system consisting of an OrbiTrap Q Exactive (Thermo Fisher Scientific, Germany) coupled to a Dionex UltiMate 3000 UHPLC system (Thermo Fisher Scientific) consisting of an UltiMate 200 SRD degas-ser, an Ultimate 3000 RS binary solvent pump system, a Di-onex UltiMate 3000 column oven and an Ultimate 3000 RS autosampler.

Chromatographic Conditions. Chromatographic separation

was achieved using an Accucore C18 2.6 µm column (2.1 × 100 mm), using gradient elution with mobile phases A and B. The gradient started with 30% B for 1 min, increased to 80% B in 0.6 min. At 2.6 min, it was set to 95% B for 2.1 min and then returned to the starting conditions to re-equilibrate the system.

The total run time was 5.5 min. The column temperature was set at 50.0°C. The flow rate was 0.500 mL/min and the injec-tion volume was 5.0 µL. The autosampler temperature was set to 12°C.

HRMS Conditions. The MS was run in full scan positive

mode with a resolving power of 17500 FWHM. The heated electrospray ionization source (HESI) was optimized as fol-lows: spray voltage 3.0 kV, capillary temperature 300°C, heater temperature 450°C, S-lens RF level 70, sheath gas flow rate 60 and auxiliary gas flow rate 18. The instrument was calibrated in positive mode using Thermo Fisher Scientific calibration solution every 48 h. A mass tolerance of 5 ppm was employed.

Method Validation

The LC-HRMS method was vali-dated according to an internal protocol for MS-based screen-ing methods. The protocol follows the rules and recommenda-tions of the FDA, ICH and Swedac.

Linrarity. The linearity of the method was determined by

analyzing 5 calibration curves prepared in 5 different plasma lots. Each calibration curve consisted of 6 calibration levels, prepared by serial 1:2 dilutions of the 1st _{calibrator (40}

ng/mL). A linear calibration model, using 1/X or 1/X2

weighting, was applied.

Detection and Quantification Limits. The limit of detection

(LOD) and lower limit of quantification (LLOQ) were deter-mined experimentally for each analyte. The LOD and LLOQ were estimated by analyzing calibrators at low concentrations, starting at 5 ng/mL and thereafter serially diluting each brator 1:2 to a final concentration of 0.3125 ng/mL. All cali-brators were spiked in pooled matrix consisting of 4 different plasma lots.

Accuracy and Precision. The precision and accuracy of the

method were determined by analyzing quality controls at 3 concentration levels; 15 ng/mL, 7.5 ng/mL and 3.75 ng/mL. Five samples at each level were analyzed on 5 different days and quantified using daily prepared calibration curves. Accu-racy was calculated and expressed in %; [(mean observed concentration)/(spiked concentration)] × 100. The precision of the method (CV %) was determined by replicates analysis (n=25) of each level. Inter- and intra-day imprecision was determined as well.

Selectivity. The selectivity of the LC-HRMS method was

evaluated by analyzing blank plasma samples from 6 different individuals.

Autosampler Stability. The stability of processed samples

was evaluated by reinjecting quality control samples that were stored in the autosampler at 12°C for 24 h. The controls were processed against the original calibration and compared using a paired t-test.

Recovery and MatrixEffect. To calculate and evaluate the

matrix effect and recovery of analytes, 3 sets of triplicates of the 10-ng/mL calibrator were prepared in 5 different plasma lots. The experiment was performed according to the recom-mendations of Matuszewski et al. [21] Set 1, prepared in

mo-bile phase A, was used to evaluate the instrument perfor-mance. The 2nd _{and 3}rd _{sets were prepared in different}

matri-ces, and samples were spiked after the precipitation procedure in Set 2 (post-extraction analysis) and before the precipitation procedure in Set 3 (pre-extraction analysis). The standard solution was prepared by diluting the final working solution (200 ng/mL) with mobile phase A to an appropriate concentra-tion.

RESULTS AND DISCUSSION

A successful attempt was made to develop a multi-component LC-HRMS screening method in serum/plasma for 43 synthetic cannabinoids. To the best of our knowledge, this is the first method utilizing protein precipitation as sample cleaning and HRMS for detection. Protein precipitation is known to be associated with major matrix effects mainly presenting as ion suppression. Therefore, to document if this was the case, ma-trix effects, recovery, and process efficiency were carefully evaluated. Accordingly, the method was demonstrated to be able to detect all analytes of interest in a qualitative manner, with an LOD range of 0.3–5.0 ng/mL.

The method was partly validated and applied on samples from authentic intoxication cases.

METHOD VALIDATION

The method was validated according to an internal protocol at the Department of Clinical Pharmacology, Karolinska Univer-sity Hospital. The protocol combines recommendations and guidelines of the ICH, FDA, and Swedac.

Method specificity was determined by analyzing six blank plasma samples and no interferences from endogenous com-pounds were observed. The detector showed excellent accura-cy, with Δm/z ranging between -0.9765 and 1.731 ppm (At-tachment 1, Table 1). Linearity, specific for each analyte, was determined in the range 40.0–0.3125 ng/mL. A 1/X or 1/X2

weighted linear regression model was applied after homogene-ity of variance was proved not to be fulfilled (Attachment 2, Table 1). The evaluation of data was based on 5 calibration curves, with 6 points each, and in 5 different plasma lots. An accepted determination coefficient (r2_{) was >0.95, with the CV}

calculated for slopes <20% (Attachment 3, Table 1). ADB-FUBINACA, EG018, URB597 and JWH210 did not fulfill the linearity requirements and should therefore be evaluated only in a qualitative manner. This problem will be further discussed in connection with the matrix effect results.

For analytes meeting the linearity criteria, the LOD and LLOQ were determined. The traditional way of LOD and LLOQ evaluation, based on signal-to-noise measurements, is not applicable using an HRMS detector, since a background noise is not present when exact masses are extracted in full scan mode. Accordingly, the LOD was determined for a concentra-tion range, when the following criteria were satisfied: scan points amount was at least 5, peak shape was satisfactory after smoothing algorithm application and dynamic range of the detector was greater than 1 × 105_{, the LOQ was determined as}

a lower concentration, and the CV did not exceed 20% (At-tachment 2, Table 1).

As mentioned before, matrix effect, recovery, and process efficiency were estimated and validated [21], using data ob-tained from 3 sets of samples prepared in 5 different matrices.

Using these 3 sets of samples, 3 types of systems can be as-sessed. By analyzing Set 1, which are neat samples in mobile phase, the total system reproducibility can be followed. The results obtained from Set 2, consisting of samples spiked after precipitation, provide good information about possible matrix effect. Comparison of variability in the CV, either increasing or decreasing, can be translated as an indication of existing ion enhancement or suppression. When analyzing differences in the variability in CV between Set 1 and 3, where samples were spiked before precipitation, a combined effect of a sample matrix effect and differences in recoveries can be observed. The matrix effect (ME), recovery (RE), and process efficiency (PE) were calculated as follows (Attachment 3, Table 1): ME (%) = Set 2/Set 1 × 100

RE (%) = Set 3/Set 2 × 100 PE (%) = Set 3/Set 1 × 100

Matrix effect was analyzed by comparing absolute peak areas for each analyte in Set 2 and 1. The values obtained in this manner (according to equation 1) are considered as an abso-lute matrix effect, that in our experiments ranged between 33% and 115% (median 66%). Most analytes showed ME values <100% which indicates ion suppression. When the ME values obtained from different plasma lots were compared (data not shown), using one-way analysis of variance (ANOVA), a significant difference between lots was observed (p<0.05). Based on these results, an assumption of the pres-ence of a relative matrix effect can be made. The wide range of ME values strongly indicated that the relative matrix effect is analyte depended and could be compensated by using inter-nal standards that undergo similar suppression or enhancement in biological samples. In multi-component methods, this ap-proach would be difficult to apply. In our case, only 2 internal standards were used for the 43 analytes, both showing an ME value near the median (Attachment 3, Table 1). To determine if ME has a significant influence on the quantification of an analyte, slopes of calibration curves and the CV were com-pared. A high CV value (>20%) would indicate a significant influence of matrix effect on an analyte and those were as-sessed only in a qualitative manner. In addition, the overall method precision should be evaluated by analyzing the QC samples at different concentrations.

Recovery and overall process efficiency were calculated for each analyte, according to equation 2 and 3. Matrix effect was taken into consideration when the recoveries were determined, by comparing data from samples spiked before precipitation with samples spiked after precipitation. Our data set showed recovery values in the range 51% and 86% (median 71%). Process efficiency was calculated by comparing absolute peak areas between samples spiked before extractions and neat samples in the mobile phase. Calculated values were between 22% and 99% (median 45%).

Quality control (QC) samples at 3 levels (High 15 ng/mL, Medium 7.5 ng/mL, and Low 3.75 ng/mL) were analyzed to evaluate overall method accuracy and precision. Five controls at each level were analyzed in 5 different batches (n=25 at each level). Quantification of the QC was performed by using daily prepared calibration curves. Accuracy was estimated for each analyte (at all levels) and was expressed by (mean ob-served concentration/spiked concentration) × 100. Precision data were presented as the relative standard deviation (RSD) and was calculated as a within-batch (repeatability), between-batch, and total precision using one-way ANOVA. [22, 23]

Accuracy data were in the range 95–120 % for all analytes at each level, except XLR11 that was 126% at the High QC level. Within-batch precision data were within the required limits of a quantitative method, with High and Medium QC ≤15% and Low ≤20, except for the following analytes: 5F-PY-PINACA (15.5% at High level and 16.5% at Medium level), 5F-ADB-PINACA (15.3% at High level), MAB-CHMINACA (20.3% at Low level), and XLR11 (19.3% at Medium level). Between-batch precision, with the same requirements, was fulfilled for most analytes, expect for 5F-ADB-PINACA (16.7 % at High level), AB-FUBINACA (16.7% at High level), CB-13 (19.5% at Low level), and PB22 (21.5% at Low level). The total precision showed greater variation and more analytes were >15%, for the High and Medium level, and >20% for the Low level. The values for total precision was in the range 5– 23% for High QC level, 6–23% for Medium level, and 5–25% for the Low level, except for MAB-CHMINACA where the RSD for the Low level was 31.7%. Compounds showing greater variation than the recommended should be considered as semi-quantitative or qualitative.

A stability test of 24-h storage in the autosampler showed no significant difference between the results of day 0 and day 1 (p>0.05, paired t-test). Therefore, no further stability tests were performed. However, it is strongly recommended to conduct long-term and freeze/thaw stability tests. It was ob-served during the experiments that the low-level calibrators of some compounds, when prepared from a stock solution stored at -20°C in 5-mL bottles, were degraded or completely lost. Accordingly, new stocks were prepared and stored in 500-µL aliquots at -80°C, and each aliquot was only used once. In this way, calibrator stability was much improved and no signifi-cant degradation was observed.

APPLICATION TO AUTHENTIC SAMPLES

A 84 patient samples were analyzed (57 males, 27 females). There were 25 samples in 2010 and 2011 and 34 in 2013. Samples originating from 38 male subjects aged 13-41, medi-an 20 medi-and 8 female subject aged 15-31, medimedi-an 20 were found positive for at least one synthetic cannabinoids. Among those 46 positive cases, 33 were evaluated as positive in toxicologi-cal screening and 13 in mass extraction analysis (without reference standards; JWH-018, JWH-081, JWH250). Positive confirmed samples are summarized in Table 1.

The following 16 parent synthetic cannabinoids were identi-fied and their structures are presented in Figure 1. The most prevalent compound was JWH-210 (13 samples), followed by JWH-081 (9 samples), 5F-AKB48 (8 samples), JWH-250 and JWH-018 (4 samples), JWH-203 and AB-FUBINACA (3 samples), 5F-PB22, AKB48 and AM2201 (2 samples), ADBICA, AB-CHMINACA, AB-PINACA, ADB-FUBINACA, FUB-AMB, MAM2201, RSC 2 and STS135 (1 sample).

A distinguishing change in sample findings was observed between different time period, Figure 2. JWH-203 was detect-ed only in samples from 2010, JWH-210 was detectdetect-ed in 2010 and 2011 together with AM2201 and AM694. JWH203 and JWH210 were put on a controlled list in 2011. AM2201 and AM694 were scheduled as narcotics, according to the Act on the Control of Narcotic Drugs, Sweden in 2012 (23). Samples from 2013 showed no trace of any compound mentioned above. Year 2013 presented a much larger variety in positive

findings; we were able to detect 12 different synthetic canna-binoids.

CONCLUSION

A LC-HRMS screening method for 43 synthetic cannabinoids in serum were developed and validated. The method is based on a simple protein precipitation. The linearity and accuracy were satisfactory for 38 analytes. Remaining five can be eval-uated in qualitative manner. The major matrix effect was ob-served for most of the analytes. Further stability experiments should be performed, as some problems were observed in this area. We also showed that positive patient samples can be detected by using only mass extraction analysis, which con-firms that this method can be keep up-to-date, with no re-quirement for reference materials.

ACKNOWLEDGMENT

I would like to thank my supervisors, professor Olof Beck, pro-fessor, Division of Clinical Pharmacology, Karolinska University Hospital and professor Anders Helander, professor, Division of Clinical Chemistry, Karolinska University Hospital for their support and encouragement during this project. I would also like to mention and thank my colleagues at Department of Clinical Pharmacology for all their help and emotional support and caring they provided.

REFERENCES

1.Martin, B.R., R. Mechoulam, and R.K. Razdan, Discovery and characterization of endogenous cannabinoids. Life Sci, 1999. 65(6-7): p. 573-95.

2.Harris, C.R. and A. Brown, Synthetic cannabinoid intoxica-tion: a case series and review. J Emerg Med, 2013. 44(2): p. 360-6.

3.Henry, H., et al., Comparison between a high-resolution single-stage Orbitrap and a triple quadrupole mass spectrome-ter for quantitative analyses of drugs. Rapid Communications in Mass Spectrometry, 2012. 26(5): p. 499-509.

4.Mackie, K., Distribution of cannabinoid receptors in the central and peripheral nervous system. Handb Exp Pharmacol, 2005(168): p. 299-325.

5.Schwartz, M.D., et al., A Common Source Outbreak of Severe Delirium Associated with Exposure to the Novel Syn-thetic Cannabinoid Adb-Pinaca. Journal of Emergency Medi-cine, 2015. 48(5): p. 573-580.

6.Van Sickle, M.D., et al., Identification and functional char-acterization of brainstem cannabinoid CB2 receptors. Science, 2005. 310(5746): p. 329-32.

7.Wiley, J.L., J.A. Marusich, and J.W. Huffman, Moving around the molecule: relationship between chemical structure and in vivo activity of synthetic cannabinoids. Life Sci, 2014. 97(1): p. 55-63.

8.Fattore, L. and W. Fratta, Beyond THC: The new generation of cannabinoid designer drugs. Front Behav Neurosci, 2011. 5: p. 60.

9.Crime, U.N.O.o.D.a., World Drug Report 2015U.N. publica-tion, Editor. 2015.

10.Isabella Gripe (red.), C.f.a.-o.n., CAN, Skolelevers drogva-nor 2015 2015: Stockholm.

11.narkotikaupplysning, C.f.a.-o., Skolelevers drogvanor 2014, A.E. (red.), Editor. 2014: Stockholm.

12.Ammann, J., et al., Detection and quantification of new designer drugs in human blood: Part 1 - Synthetic canna-binoids. J Anal Toxicol, 2012. 36(6): p. 372-80.

13.Castaneto, M.S., et al., Quantitative urine confirmatory testing for synthetic cannabinoids in randomly collected urine specimens. Drug Test Anal, 2015. 7(6): p. 483-93.

14.Dresen, S., et al., Development and validation of a liquid chromatography-tandem mass spectrometry method for the quantitation of synthetic cannabinoids of the aminoalkylindole type and methanandamide in serum and its application to forensic samples. Journal of Mass Spectrometry, 2011. 46(2): p. 163-171.

15.Grabenauer, M., et al., Analysis of Synthetic Cannabinoids Using High-Resolution Mass Spectrometry and Mass Defect Filtering: Implications for Nontargeted Screening of Designer Drugs. Analytical Chemistry, 2012. 84(13): p. 5574-5581. 16.Hudson, S., et al., Use of High-Resolution Accurate Mass Spectrometry to Detect Reported and Previously Unreported Cannabinomimetics in "Herbal High" Products. Journal of Analytical Toxicology, 2010. 34(5): p. 252-260.

17.Kneisel, S. and V. Auwarter, Analysis of 30 synthetic can-nabinoids in serum by liquid chromatography-electrospray ionization tandem mass spectrometry after liquid-liquid ex-traction. J Mass Spectrom, 2012. 47(7): p. 825-35.

18.Kronstrand, R., et al., LC-QTOF-MS as a superior strategy to immunoassay for the comprehensive analysis of synthetic cannabinoids in urine. Analytical and Bioanalytical Chemistry, 2014. 406(15): p. 3599-3609.

19.Scheidweiler, K.B., M.J.Y. Jarvis, and M.A. Huestis, Non-targeted SWATH acquisition for identifying 47 synthetic cannabinoid metabolites in human urine by liquid chromatog-raphy-high-resolution tandem mass spectrometry. Analytical and Bioanalytical Chemistry, 2015. 407(3): p. 883-897. 20.Wu, A.H., et al., Role of liquid chromatography-high-resolution mass spectrometry (LC-HR/MS) in clinical toxicol-ogy. Clinical Toxicology, 2012. 50(8): p. 733-742.

21.Matuszewski, B.K., M.L. Constanzer, and C.M. Chavez-Eng, Strategies for the assessment of matrix effect in quantita-tive bioanalytical methods based on HPLC-MS/MS. Analyti-cal Chemistry, 2003. 75(13): p. 3019-3030.

22.Beyer, J., et al., Detection and validated quantification of nine herbal phenalkylamines and methcathinone in human blood plasma by LC-MS/MS with electrospray ionization. J Mass Spectrom, 2007. 42(2): p. 150-60.

23.Massart, D.L., et al., Handbook of chemometrics and qualimetrics: Part A. 1997: Elsevier Science Inc.

8

Table1. Summary of positive founding in authentic case samples.

Age Gender Toxicological

finding Concentration (ng/mL) Mass extrac-tion analysis Brand name of Spice-mix Reported Synthetic cannabinoids 20 female 5F-AKB48 6.64 - - - 30 female 5F-AKB48 3.74 - - - 24 male 5F-AKB48 2.18 - - - 18 male 5F-AKB48 14.21 - - - 37 female 5F-AKB48 5F-PB22 AKB48 20.85 4.26 2.5 - - 5F-AKB48 PB22 26 male 5F-AKB48 AB-FUBINACA FUB-AMB 3.74 X1 0.87 - - - 18 male 5F-AKB48 STS135 7.77 15.18 - - PB22 STS135 18 male 5F-PB22 2.65 - - 5F-PB22 23 male AB-CHMINACA 3.86 - - - 16 male AB-FUBINACA X - - - 26 male AB-FUBINACA X - - - 17 male AB-PINACA ADB-FUBINACA ADBICA <2.5 X 30.25 - Vevivs - 24 male AKB-48 5F-AKB48 8.59 <2.5 - - - 17 female AM2201 <0.6 - - - 24 male AM2201 <0.6 - - - 15 male JWH-203 <1.25 - Diablo - 19 male JWH-203 12.43 - Remix - 24 male JWH-203 3.18 - - - 28 male JWH-210 X - - - 30 male JWH-210 X - - - 15 male JWH-210 X - - JWH-015 18 male JWH-210 X - Bonzai - 17 male JWH-210 X - Boom - 15 female JWH-210 X - - - 13 male JWH-210 X - - -

24 male JWH-210 X - Jamaican gold

extreme

JWH-210

19 male JWH-210 X - Jamaican gold

extreme

JWH-210

15 male JWH-210 X - Jamaican gold

extreme

JWH-210

20 female JWH-210 X - Wasted JWH-210

9

26 female JWH-210 X - Jamaican gold

extreme JWH-210 41 male JWH-210 X - - JWH-210 17 male MAM2201 >40 - - - 16 male - - JWH-081 - - 20 male - - JWH-018 Bonzai JWH-018

20 male - - JWH-081 Jamaican gold

extreme JWH-081 25 male - - JWH-081 - JWH-081 15 male - - JWH-081 - JWH-081 25 male - - JWH-081 JWH-250 Aromatic potpourri JWH-081 JWH-250 17 male - - JWH-081 JWH-250 - JWH-081 21 male - - JWH-081 JWH-250 - JWH-081 20 male - - JWH-250 - JWH-081 JWH-250

21 male - - JWH-081 Jamaican gold

extreme JWH-081 21 male - - JWH-081 DJ JWH-081 18 male - - JWH-018 - JWH-018 17 female - - JWH-018 - - 29 male RSC2 2.46 JWH-018 Finfedron -

10

Figure 1. Summary of synthetic cannabinoids that were found in positive screened samples. 16 different compounds were identified. The most prevalent compound was JWH-210 (13 samples), followed by JWH-081 (9 samples), 5F-AKB48 (8 samples), JWH-250 and JWH-018 (4 samples), JWH-203 and AB-FUBINACA (3 samples), 5F-PB22, AKB48 and AM2201 (2 samples), ADBICA, AB-CHMINACA, AB-PINACA, ADB-FUBINACA, FUB-AMB, MAM2201, RSC 2 and STS135 (1 sample) 0 2 4 6 8 10 12 14 Ca se n umb er

Synthetic cannabinoids identified in patien

samples

11

Figure 2 Synthetic cannabinoids detected in authentic case samples in 2010, 2011 and 2013. A characteristic change in sam-ple findings was observed between different time period. JWH-203 was detected only in samsam-ples from 2010, JWH-210 was detected in 2010 and 2011 together with AM2201 and AM694. JWH203 and JWH210 were put on a controlled list in 2011. AM2201 and AM694 were scheduled as narcotics. Samples from 2013 showed no trace of any JWH-203, JWH-210, AM2201 or AM694. 0 1 2 3 4 5 6 7 8 Case number

Synthetic cannabinoids detected in patient samples in

years 2010, 2011 nad 2013

12

ATTACHMENT 1

Table 1 Molecular formula, theoretical and measured mass of protonated compound, mass error (Δm/z, ppm) and reten-tion time (RT, min) for 43 synthetic cannabinoids

Compound Molecular

for-mula Theoretical mass [M+H]+ Measured mass [M+H]+ Δm/z (ppm) RT (min) Internal Standard

5PY-PINACA C17H22FN3O 304.18197 304.18213 0.52241 2.49 AM2201 d5 5F-ADB PINACA C19H27FN4O2 363.21908 363.21912 0.09969 2.46 AM2201 d5 5F-AKB48 C23H30FN3O 384.24457 384.24496 1.02695 3.32 JWH210 d9 5F-AMB C19H26FN3O3 364.2031 364.20285 -0.68549 2.62 AM2201 d5 5F-PB22 C23H21FN2O2 377.16598 377.16611 0.33719 2.64 AM2201 d5 A834735 C22H29NO2 340.22711 340.22717 0.18473 2.84 AM2201 d5 AB-CHMINACA C20H28N4O2 357.2285 357.22864 0.38545 2.79 AM2201 d5 AB-FUBINACA C20H21FN4O2 369.17213 369.17227 0.38391 2.42 AM2201 d5 AB-PINACA C18H26N4O2 331.21285 331.21283 -0.06162 2.62 AM2201 d5

AB001 C24H31NO 350.24784 350.24777 -0.19353 3.77 JWH210 d9

ADB-FUBINACA C21H23FN4O2 383.18778 383.18805 0.70283 2.54 AM2201 d5 ADBICA C20H29N3O2 344.23325 344.23352 0.78583 2.75 AM2201 d5 AKB48 C23H31N3O 366.25399 366.25421 0.60457 3.92 JWH210 d9

AM1220 C26H26N2O 383.21179 383.21194 0.40338 2.38 AM2201 d5

AM2201 C24H22FNO 360.17582 360.17581 -0.02286 2.80 AM2201 d5 AM2233 C22H23IN2O 459.09278 459.09305 0.58397 2.25 AM2201 d5 AM694 C20H19FINO 436.05681 436.05713 0.73134 2.63 AM2201 d5 BB22 C25H24N2O2 385.19105 385.19067 -0.97659 3.17 JWH210 d9

CB-13 C26H24O2 369.18491 369.18546 1.47710 4.30 JWH210 d9

EG018 C28H25NO 392.20089 392.20126 0.95213 4.15 JWH210 d9

EMB-FUBINACA C22H24FN3O3 398.18745 398.18756 0.27885 2.84 AM2201 d5 FUB-AMB C21H22FN3O3 384.1718 384.17184 0.11579 2.70 AM2201 d5

JWH-015 C23H21NO 328.16959 328.16965 0.17435 2.89 AM2201 d5 JWH-020 C26H27NO 370.21654 370.21667 0.36412 3.61 JWH210 d9 JWH-098 C26H27NO2 386.21146 386.21161 0.38551 3.33 JWH210 d9 JWH-147 C27H27NO 382.21654 382.21664 0.27285 3.61 JWH210 d9 JWH-203 C21H22ClNO 340.14627 340.14618 -0.26695 3.11 JWH210 d9 JWH-210 C26H27NO 370.21654 370.21701 1.27087 3.52 JWH210 d9 JWH-251 C22H25NO 320.20089 320.20102 0.40377 3.10 JWH210 d9 JWH-307 C26H24FNO 386.19147 386.19214 1.73145 3.32 JWH210 d9 JWH-398 C24H22ClNO 376.14627 376.14664 0.97559 3.54 JWH210 d9 MAB-CHMINACA C21H30N4O2 371.24415 371.24445 0.79678 2.95 AM2201 d5 MAM-2201 C25H24FNO 374.19147 374.19147 -0.00726 2.95 AM2201 d5 MDMB-CHMICA C23H32N2O3 385.24857 385.24866 0.22642 3.18 JWH210 d9 MMB-2201 C20H27FN2O3 363.20785 363.20792 0.18243 2.52 AM2201 d5 MN-18 C23H23N3O 358.19139 358.19180 1.15295 3.39 JWH210 d9 PB-22 C23H22N2O2 359.1754 359.17532 -0.21303 2.97 AM2201 d5 RSC-2 C21H23NO2 322.18016 322.18027 0.33315 2.82 AM2201 d5 STS-135 C24H31FN2O 383.24932 383.24951 0.50025 3.12 JWH210 d9 THJ-018 C23H22N2O 343.18049 343.18094 1.30754 3.39 JWH210 d9 UR-144 C21H29NO 312.23219 312.23248 0.93812 3.41 JWH210 d9

URB-597 C20H22N2O3 339.17032 339.17035 0.08586 2.37 AM2201 d5

13

ATTACHMENT 2

Table 1 Linearity, LOD and LOQ results.

Compound Linearity (ng/mL) Weighting R2 + - SD LOD LOQ Internal Standard

5PY-PINACA 40-0.3 1/X2 _{0.9977 + 0.0016 <0.3 0.3 AM2201 d5}

5F-ADB PINACA 40-0.3 1/X2 _{0.9967 + 0.0029 <0.3 0.3 AM2201 d5}

5F-AKB48 40-1.25 1/X 0.9966 + 0.0020 1.25-0.6 1.25 JWH210 d9 5F-AMB 40-1.25 1/X2 _{0.9979 + 0.0020 <0.3 1.25 AM2201 d5} 5F-PB22 40-0.6 1/X2 _{0.9981+ 0.0018 0.6-0.3 0.6 AM2201 d5} A834735 40-2.5 1/X 0.9966 + 0.0028 2.5-1.25 2.5 AM2201 d5 AB-CHMINACA 40-5 1/X2 _{0.9987 + 0.0008 5-2.5 5 AM2201 d5} AB-FUBINACA X2 _{X X X X AM2201 d5} AB-PINACA 40-1.25 1/X 0.9980 + 0.0011 1.25-0.6 1.25 AM2201 d5 AB001 40-2.5 1/X2 _{0.9979 + 0.0026 1.25-0.6 2.5 JWH210 d9} ADB-FUBINACA X 1/X2 _{0.9973 + 0.0022 5-2.5 X AM2201 d5} ADBICA 40-5 1/X2 _{0.9982 + 0.0011 5-2.5 5 AM2201 d5} AKB48 40-5 1/X2 _{0.9990 + 0.0007 2.5-1.25 5 JWH210 d9} AM1220 40-0.6 1/X2 _{0.9985 + 0.0009 <0.3 0.6 AM2201 d5} AM2201 40-1.25 1/X2 _{0.9990 + 0.0004 1.25-0.6 1.25 AM2201 d5} AM2233 40-0.6 1/X2 _{0.9974 + 0.0025 <0.3 0.6 AM2201 d5} AM694 40-1.25 1/X2 _{0.9984 + 0.0009 <0.3 1.25 AM2201 d5} BB22 40-2.5 1/X2 _{0.9982 + 0.0021 2.5-1.25 2.5 JWH210 d9} CB-13 40-5 1/X2 _{0.9930 + 0.0050 2.5-1.25 5 JWH210 d9} EG018 X X X X X JWH210 d9 EMB-FUBINACA 40-1.25 1/X2 _{0.9932 + 0.0105 1.25-0.6 1.25 AM2201 d5} FUB-AMB 40-1.25 1/X2 _{0.9980 + 0.0013 <0.3 1.25 AM2201 d5} JWH-015 40-0.3 1/X2 _{0.9769 + 0.0311 <0.3 0.3 AM2201 d5} JWH-020 40-5 1/X2 _{0.9973 + 0.0034 2.5-1.25 5 JWH210 d9} JWH-098 40-1.25 1/X2 _{0.9974 + 0.0036 1.25-0.6 1.25 JWH210 d9} JWH-147 40-2.5 1/X2 _{0.9952 + 0.0080 1.25-0.6 2.5 JWH210 d9} JWH-203 40-1.25 1/X2 _{0.9950 + 0.0039 1.25-0.6 1.25 JWH210 d9} JWH-210 X X X X X JWH210 d9 JWH-251 40-5 1/X2 _{0.9935 + 0.0079 <0.3 5 JWH210 d9} JWH-307 40-5 1/X 0.9976 + 0.0027 <0.3 5 JWH210 d9 JWH-398 40-5 1/X2 _{0.9980 + 0.0014 5 5 JWH210 d9} MAB-CHMINACA 40-2.5 1/X2 _{0.9985 + 0.0009 2.5-1.25 2.5 AM2201 d5} MAM-2201 40-1.25 1/X2 _{0.9803 + 0.0251 0.6-0.3 1.25 AM2201 d5} MDMB-CHMICA 40-2.5 1/X2 _{0.9985 + 0.0008 <0.3 2.5 JWH210 d9} MMB-2201 40-1.25 1/X2 _{0.9926 + 0.0107 <0.3 1.25 AM2201 d5} MN-18 40-5 1/X2 _{0.9963 + 0.0059 2.5-1.25 5 JWH210 d9} PB-22 40-1.25 1/X2 _{0.9950 + 0.0056 <0.3 1.25 AM2201 d5} RSC-2 40-1.25 1/X 0.9958 + 0.0037 0.6-0.3 1.25 AM2201 d5 STS-135 40-0.6 1/X2 _{0.9971 + 0.0015 0.6-0.3 0.6 JWH210 d9} THJ-018 40-5 1/X2 _{0.9973 + 0.0025 2.5-1.25 5 JWH210 d9} UR-144 40-2.5 1/X2 _{0.9972 + 0.0019 <0.3 2.5 JWH210 d9} URB-597 X X X X X AM2201 d5 XLR-11 20-5 1/X2 _{0.9796 + 0.0323 <0.3 5 JWH210 d9}

14

ATTACHMENT 3

Table 1 Recovery (RE), Matrix Effect (ME), Process Efficiency (PE) results. Calculations were based on data obtained from five different plasma lots. Mean slope and coefficient of variance (%) of the standard curves in five different plasma lot of samples spiked before precipitation.

Compound RE (%) ME (%) PE (%) Slopesa _{CV (%) slopes}

5PY-PINACA 71% 82% 58% 1,19E+02 17%

5F-ADB PINACA 78% 87% 68% 1,42E+01 15%

5F-AKB48 68% 67% 46% 2,10E+02 4% 5F-AMB 73% 73% 53% 5,50E+01 11% 5F-PB22 69% 66% 46% 2,80E+01 3% A834735 51% 60% 31% 4,23E+01 5% AB-CHMINACA 63% 96% 61% 9,61E+00 6% AB-FUBINACA 86% 115% 99% 7,66E+00 11% AB-PINACA 78% 95% 74% 1,31E+01 16% AB001 69% 56% 38% 5,04E+01 6% ADB-FUBINACA 82% 57% 47% 4,00E+00 41% ADBICA 74% 99% 73% 8,68E+00 7% AKB48 63% 68% 43% 5,90E+01 7% AM1220 68% 64% 43% 5,98E+01 7% AM2201 59% 62% 36% 4,83E+01 3% AM2233 64% 66% 42% 5,41E+01 9% AM694 72% 58% 41% 4,48E+01 6% BB22 71% 66% 47% 1,26E+02 7% CB-13 69% 68% 47% 1,69E+01 12% EG018 62% 67% 42% 1,17E+01 37% EMB-FUBINACA 57% 72% 41% 4,44E+01 4% FUB-AMB 55% 76% 41% 4,96E+01 5% JWH-015 69% 71% 49% 6,40E+01 4% JWH-020 61% 39% 24% 4,94E+01 3% JWH-098 77% 67% 51% 2,13E+02 6% JWH-147 83% 45% 37% 4,11E+01 2% JWH-203 71% 62% 44% 2,07E+02 10% JWH-210 78% 67% 52% 4,63E+01 23% JWH-251 76% 73% 55% 3,73E+02 10% JWH-307 76% 55% 42% 1,36E+02 13% JWH-398 81% 57% 46% 2,61E+01 7% MAB-CHMINACA 77% 74% 57% 9,18E+00 3% MAM-2201 71% 61% 43% 1,33E+01 4% MDMB-CHMICA 78% 51% 40% 2,01E+02 9% MMB-2201 67% 52% 35% 3,23E+01 19% MN-18 70% 33% 23% 1,42E+02 7% PB-22 72% 86% 62% 1,64E+01 5% RSC-2 71% 66% 47% 6,13E+01 2% STS-135 72% 61% 44% 3,63E+02 11% THJ-018 73% 38% 28% 1,27E+02 4% UR-144 58% 66% 38% 2,03E+02 5% URB-597 72% 87% 63% 1,21E+00 42% XLR-11 73% 59% 43% 4,84E+02 13% AM2201-d5 64% 66% 42% x x

15

16

ATTACHMENT 4

Table 1 Accuracy. Intraday (results presented for 1 run), interday and total imprecision (%). High QC 15 ng/mL, medium QC 7.5 ng/mL and low 3.75 ng/mL.

Compound Intraday imprecision

(CV %)

Interday imprecision (CV %)

Total imprecision (CV %) Accuracy (Bias %)

High Med Low High Med Low High Med Low High Med Low

5F-PY-PINACA 15,50 16,46 13,04 11,08 9,36 8,33 19,06 18,93 15,47 109,1 103,8 110,3 5F-ADB PINACA 15,29 14,88 14,27 16,73 14,29 10,80 22,67 20,63 17,90 107,0 98,3 105,8 5F-AKB48 11,08 9,73 9,84 5,60 2,72 7,56 12,41 10,10 12,41 114,0 114,1 113,6 5F-AMB 11,22 11,22 8,94 10,93 5,22 10,15 15,67 12,37 13,53 108,7 99,7 105,4 5F-PB22 4,77 7,46 9,40 5,88 4,50 3,96 7,57 8,71 10,20 98,7 94,7 98,2 A834735 7,89 7,70 12,40 4,78 0,00 10,98 9,22 7,45 16,56 119,4 109,5 98,5 AB-CHMINACA 7,47 12,32 x 7,95 9,79 x 10,90 15,74 x 118,5 111,1 x AB-FUBINACA 14,17 13,41 11,96 16,71 16,32 9,63 21,91 21,12 15,36 112,6 105,7 114,5 AB-PINACA 10,97 8,87 14,92 6,05 7,80 11,14 12,53 11,81 18,62 103,1 97,8 102,0 AB001 7,14 10,29 11,22 11,38 3,58 7,50 13,43 10,90 13,49 112,6 104,3 105,3 ADB-FUBINACA x3 _{x x x x x x x x x x x} ADBICA 7,86 11,84 x 9,77 8,72 x 12,54 14,71 x 104,9 98,8 x AKB48 7,96 10,95 x 13,69 4,54 x 15,84 11,85 x 105,4 103,8 x AM1220 7,96 11,16 7,33 13,69 10,74 4,49 15,84 15,49 8,60 110,3 106,4 113,3 AM2201 5,16 8,27 8,24 0,00 0,00 0,00 5,12 8,01 8,04 113,3 110,4 111,8 AM2233 9,93 12,30 9,12 11,74 6,87 6,05 15,37 14,09 10,94 110,9 105,4 114,5 AM694 4,40 6,88 10,34 9,21 4,65 7,15 10,21 8,30 12,57 103,5 95,8 99,1 BB22 8,59 11,46 11,19 6,93 0,00 5,82 11,04 11,03 12,62 112,3 108,4 102,3 CB-13 6,90 8,54 x 19,48 11,52 x 20,66 14,34 x 110,8 110,9 x EG018 x x x x x x x x x x x x EMB-FUBINACA 6,57 7,12 12,26 2,03 3,27 0,00 6,88 7,84 11,20 114,7 109,1 109,9 FUB-AMB 6,30 7,44 7,04 5,17 4,16 6,78 8,15 8,52 9,77 110,4 102,3 103,1 JWH-015 5,74 12,97 7,54 6,87 7,79 2,08 8,95 15,13 7,82 112,7 111,9 113,4 JWH-020 9,02 9,59 x 3,88 0,00 x 9,81 9,15 x 107,5 108,5 x JWH-098 10,21 10,76 12,44 9,90 0,00 3,97 14,22 10,57 13,06 110,6 110,4 110,8 JWH-147 8,02 7,23 11,13 9,70 0,00 17,41 12,59 6,82 20,67 109,9 107,8 104,8 JWH-203 9,61 8,19 12,27 10,31 6,90 0,00 14,09 10,71 11,98 118,2 118,0 114,9 JWH-210 8,95 14,83 14,78 7,14 0,00 9,99 11,45 14,30 17,84 114,4 106,4 94,0 JWH-251 8,67 7,70 x 11,42 7,44 x 14,34 10,70 x 114,4 114,2 x JWH-307 7,95 9,67 x 7,34 6,45 x 10,82 11,62 x 111,2 109,1 x JWH-398 13,27 14,73 x 7,30 8,98 x 15,14 17,25 x 120,3 104,3 x MAB-CHMINACA 10,80 8,89 20,30 14,98 8,49 24,30 18,46 12,30 31,67 120,0 115,1 94,4 MAM-2201 7,31 6,33 13,72 9,04 12,14 5,82 11,63 13,69 14,91 114,4 113,1 113,0 MDMB-CHMICA 9,59 7,28 9,04 12,65 6,70 5,17 15,87 9,90 10,41 117,4 118,2 114,8 MMB-2201 12,51 14,55 12,48 11,27 5,26 8,50 16,84 15,47 15,10 116,0 104,8 111,3 MN-18 7,33 7,80 x 7,62 4,80 x 10,58 9,16 x 107,8 109,9 x PB-22 7,38 8,94 14,42 10,32 12,02 21,45 12,68 14,98 25,84 108,8 107,6 105,0 RSC-2 5,36 5,12 3,99 4,25 5,92 3,63 6,84 7,83 5,39 111,6 107,0 110,8 STS-135 9,57 6,90 10,34 10,05 6,80 5,58 13,88 9,69 11,75 110,4 112,5 112,9

17

THJ-018 8,14 8,11 x 11,11 3,52 x 13,77 8,84 x 108,4 107,3 106,1

UR-144 8,34 8,34 8,34 12,87 5,18 5,18 15,34 9,82 9,82 114,4 109,8 112,0

URB-597 x x x x x x x x x x x x