Development and Validation of a Single Step GC/MS Method for the Determination of 41 Drugs and Drugs of Abuse in Postmortem Blood

Abstract

A toxicology laboratory often receives a high number of samples from cases (autopsies or clinical) that may require the quick delivery of trustworthy, accurate results. Thus, there is a great need for a fast and reliable method that is capable of identifying and determining a large number of drugs and drugs of abuse in biological matrices, and especially in blood. In the present study, we describe the development of a fast and simple gas chromatography–mass spectrometry (GC-MS) method for the determination of 41 drugs and drugs of abuse (DOA) in blood. Sample pre-treatment by alkaline liquid–liquid extraction (LLE) was studied through the utilization of different solvents and solvent-to-sample ratios (v/v), which aimed to achieve a greater extraction efficiency and detection sensitivity with a decreased need for large sample volumes. Butyl acetate with a sample-to-solvent ratio of 4:1 (1 mL blood: 0.25 mL butyl acetate) was the most efficient. The method was validated for all analytes, and the evaluation parameters were within the acceptance criteria. The coefficient of determination (R2) was between 0.9934 and 1, the limits of detection (LODs) ranged between 1 ng/mL and 113 ng/mL, and the limits of quantification (LOQs) were between 4 ng/mL and 375 ng/mL for all analytes. The determinations were accurate (accuracy% from 84% to 114%) and precise (RSD% from 0.66% to 14.8% for low concentrations). Deconvolution Reporting Software (DRS) for GC-MS was optimized and applied for data analysis to enhance the identification potential, thereby avoiding false identifications (false positives) and increased productivity. The NIST Automated Mass Spectral Deconvolution and Identification Software (AMDIS) and the analytical utility Retention Time Lock (RTL) Database Library assisted in data evaluation. The method was applied to 89 postmortem cases (history of mental disorders and use of psychiatric pharmaceuticals) in which diazepam (0.13 to 4.34 μg/mL), citalopram (0.04 to 0.24 μg/mL), alprazolam (0.01 to 0.12 μg/mL), olanzapine (0.009 to 0.083 μg/mL), mirtazapine (0.01 to 0.33 μg/mL), venlafaxine (0.006 to 0.92 μg/mL), haloperidol (0.007 to 0.13 μg/mL), and zolpidem (0.01 to 0.16 μg/mL) were successfully quantitated.

1. Introduction

In the framework of toxicological analysis, the main groups of drugs and drugs of abuse (DOA) that have to be investigated include anesthetics, benzodiazepines, antipsychotics, antiepileptics, opiates, cocaine, cannabinoids, and amphetamines [1]. Most of them have therapeutic applications, while some are used almost exclusively by people addicted to drugs.

In the literature, there is a large number of reported methods for the determination of pharmaceuticals and drugs in biological fluids [2,3,4,5,6,7,8,9,10,11], and especially in blood [12,13,14,15,16,17,18,19], for the purposes of a toxicology lab, while the routinely applied methods include either immunochemical methods or methods based on hyphenated techniques, such as GC-MS and liquid chromatography–mass spectrometry (LC-MS). Most of them focus on the identification of specific categories of drugs, such as opiates, benzodiazepines, etc. Immunochemical methods provide quick results, as they do not require sample pretreatment, but confirmation with a sensitive and selective chromatographic technique is still required; a false positive result due to cross reactivity with other compounds is possible, and only qualitative results can be acquired [20,21]. Furthermore, the analysis of different groups of analytes necessitates the use of different immunoassay kits, a fact that increases the cost and complexity of the analytical process.

The scientific community takes efforts to develop methods able to detect and quantitate the largest possible number of medicines and drugs. The major aim is the development of a quick, effective, and sensitive method that would replace immunochemical analysis. The most common methods applied are based on LC-MS/MS or LC-TOF/MS instrumentation due to the large number of compounds that can be easily quantitated without time-consuming sample preparation; usually only protein precipitation is needed [22,23,24]. The main drawback of these techniques is the limitation in identifying unknown compounds, as commercial mass spectral libraries are still under development and only in-house libraries can aid. Gas chromatography coupled with mass spectrometry (GC-MS) is probably the most widely accepted technique in forensic toxicology, as it has the advantage of offering drug identification based on the extended libraries that exist and, at the same time, it can provide sensitive and accurate determinations. Thus, GC-MS-based methods are considered highly reliable for the detection and determination of drugs in biological samples, and specifically in blood. The most notable studies of the last decade that focus on the determination of drugs in biological fluids by GC-MS all have one common feature that concerns the sample preparation step. Either solid phase extraction [25,26,27,28] or liquid–liquid extraction [29,30,31] is applied, and they all require an evaporation step and the majority of them make use of large volumes of organic solvents. In addition, in almost all cases, the evaporation step is followed by a derivatization step in order to increase sensitivity [32,33,34,35,36,37]. Recently, QuEChERS (Quick Easy Cheap Effective Rugged Safe) methodology has also been applied for toxicological analysis purposes, however evaporation is still required [38,39].

Here, we report the development, validation, and application of a GC-MS method for the separation and determination of 41 pharmaceuticals and drugs from different classes, which includes a single and quick sample preparation step, thereby avoiding the evaporation and derivatization processes. The developed method has been successfully applied in a large number of samples and interchangeably for the investigation of several cases of clinical and forensic interest in the Laboratory of Forensic Toxicology of the University of Thessaloniki and the Laboratory of Forensic Service of Ministry of Justice of Thessaloniki; the results obtained in each laboratory were comparable, thus it has been proven a valuable tool that greatly supports and advances the routine analysis of the two forensic laboratories.

2. Experimental

2.1. Reagents and Instrumental Conditions

All solvents were of HPLC grade. Methanol was obtained from Fisher Scientific, (Leicestershire, UK); and butyl acetate, hexane, isooctane, 4-methyl-2-pentanone, and 1-butanol were purchased from Sigma–Aldrich (St. Louis, MO, USA). Ethyl acetate was supplied from Penta (Livingston, NJ, USA). All the reference standards had more than 98.5% purity and were purchased from Lipomed AG (Arlesheim, Switzerland), except for amantadine, moclobemide, propofol, ropivacaine, biperiden, olanzapine, haloperidol, and nordazepam-d5 (Internal Standard), which were from LGC Standards (Teddington, UK). Human blood samples were obtained after informed consent from healthy donors and were screened by GC-MS for the presence of the investigated drugs and DOA.

Analysis by GC-MS was performed on an Agilent Technologies 7890A GC, combined with a 5975C inert XL EI/CI MSD with a Triple-Axis Detector (Agilent Technologies, Santa Clara, CA, USA) equipped with a CTC autosampler. The method had a total duration of 21 min with the following temperature program: Initially, 120 °C for 1 min, then increased to 300 °C with a rate of 15 °C/min, and followed by a backflush program at 300 °C for 10 min. The injection of 1 μL of sample was performed through a PTV injector operating between 200 °C to 320 °C. The mass spectrometer (MS) was operated at electron impact ionization mode (EI, 70 eV) in full scan mode at a mass range from 40 to 500 amu. The initial pressure for this instrument was 19.418 psi with a set flow of 1.2 mL/min. The second GC-MS system used was also an Agilent Technologies 7890A GC, combined with an 5975C inrtXL EI/CI MSD with a Triple-Axis Detector (Agilent Technologies, Santa Clara, CA, USA) equipped with an Agilent G4513A autosampler. The GC-MS analysis employed an Optima-5-ms, 30 m × 250 μm × 0.25 μm (Scitech Scientific, West Perth, Western Australia) column. The method had a total duration of 31 min with the following temperature program: Initially, 120 °C and increasing to 300 °C (15 °C/min). An injection of 1 μL of sample was done through a split-splitless injector operating at 300 °C. The mass spectrometer (MS) was operated at electron impact ionization mode (EI, 70 eV) and the collected mass range was 40–500 amu. The applied flow was also 1.2 mL/min for this instrument in order to achieve common retention times with an initial pressure at 14.012 psi. On both laboratories GC separations, a 30 m Agilent J&W HP-5ms capillary column with a film thickness of 0.25 μm and an i.d. of 0.25 μm were performed. Backflash was performed with a 1.5 m deactivated Agilent column with a film thickness of 0.18 mm. Both instruments also had the same MS source temperature (230 °C) and MS quadrupole temperature (150 °C).

For the identification and quantification of the analytes, a full scan mode was selected; the confirmation ions and the analyte’s retention times (RT) are presented in

Table 1. Retention times and target ions for the 41 drugs and the IS.

No	Compound	RT (min)	Target Ion	Qualifier Ions
1	Methamphetamine	3.18	58	91	65
2	Amantadine	3.69	94	151	108
3	Propofol	4.45	163	178	117
4	MDA	5.51	44	135	136
5	MDMA	5.96	58	135	77
6	MDEA	6.22	72	44	135
7	Bupropion	6.46	44	100	57
8	MBDB	6.65	72	135	89
9	Fluoxetine	8.42	44	104	162
10	Ketamine	8.49	209	180	182
11	Lidocaine	8.53	58	86	120
12	Tramadol	9.10	58	263	135
13	Phenobarbital	9.25	204	232	115
14	Venlafaxine	9.99	58	134	179
15	Methadone	10.33	72	294	309
16	Ropivacaine	10.71	126	84	127
17	Amitriptyline	10.73	58	275	30
18	Cocaine	10.77	82	182	303
19	Atropine	10.81	124	289	140
20	Nortriptyline	10.87	44	202	189
21	Moclobemide	11.13	100	139	113
22	Mirtazapine	11.14	195	208	180
23	Biperiden	11.29	98	218	55
24	Phenytoin	11.63	180	104	77
25	Sertaline	11.80	274	262	304
26	Citalopram	11.96	58	324	238
27	Codeine	11.97	299	162	115
28	Clomipramine	12.02	58	85	268
29	Diazepam	12.21	283	256	221
30	Chlorpromazine	12.56	58	318	86
31	Nordazepam	12.61	270	242	269
32	Midazolam	13.04	310	325	163
33	Flunitrazepam	13.14	312	285	266
34	7-AF	13.33	283	255	254
35	Fentanyl	13.72	245	146	189
36	Olanzapine	13.88	242	229	213
37	Zolpidem	14.26	235	307	219
38	Clozapine	14.88	243	256	192
39	Haloperidol	15.56	224	237	42
40	Alprazolam	15.47	204	279	308
41	Quetiapine	18.45	210	239	321
IS	Nordazepam-D5	12.61	275	247	274

with the quantification ion for each one shown in bold.

2.2. Preparation of Standard Solutions and QCs

Stock solutions of all analytes at 1.0 mg/mL were prepared in methanol. Working solutions were prepared from stock by dilution with methanol. Drug-free blood samples were used for all method optimization and validation assays. For calibration standards and quality controls (QCs), 100 μL of the working standards were added to 1 mL of whole blood after the addition of 25 μL of internal standard (IS). All solutions and QC samples were stored at −20 °C. Three groups of mixtures were prepared according to the studied concentration levels.

In order to study the selectivity of the method, blood spiked with drugs and pharmaceuticals that could be found in real samples were tested for interferences in the determination of the 41 analytes. Thus, blood was spiked at concentrations of 0.125 μg/mL for flunitrazepam and fentanyl; at 1.25 μg/mL for methamphetamine, propofol, MDA, MDMA, ecgonine methyl ester, MDEA, MBDB, ketamine, lidocaine, methadone, amitryptiline, nortriptyline, cocaine, atropine, sertraline, codeine, chlomipramine, diazepam, clozapine, and haloperidol; at 0.4 μg/mL for chlorpromazine and ropivacaine; at 0.5 μg/mL for bromazepam, nordazepam, citalopram, bupropion, fluoxetine, tramadol, mirtazapine, biperiden, nordazepam-D5 (IS), midazolam, olanzapine, zolpidem, and alprazolam; and at 5.0 μg/mL for phenobarbital, carbamazepine, phenytoin, and amantadine.

2.3. Biological Samples

The blood samples were collected at the Forensic Service of Greek Ministry of Justice in Thessaloniki during autopsy 14 to 20 h postmortem from subjects with a history of pharmaceutical use for mental disorders. The obtained samples were screened in the frame of the routine toxicological analysis and were analysed by the developed method for quantification.

2.4. Sample Preparation

To select the most efficient LLΕ protocol, the extraction recoveries of the analytes were evaluated using six different commonly used extraction solvents. More specifically, butyl acetate, ethyl acetate, hexane, isooctane, 4-methyl-2-pentanone, and 1-butanol were selected. The sample-to-solvent ratio was also evaluated in terms of recovery for the optimum solvent. Recovery was assessed by the percentage peak ratio of the peak area of the analyte that was spiked before the extraction to the peak area of the standard solution of the analyte at the same concentration. Finally, the selected applied protocol was as follows:

In 1 mL of blood sample, 25 μL of nordazepam-D5 (IS) solution (10 μg/mL) and 100 μL of the spiking solution were added. Then, the sample was alkalized by adding 500 μL of a saturated aqueous solution of K₂CO₃ (pH = 12). Finally, 250 μL of butyl acetate were added and the sample was shaken for two min following centrifugation at 6000× g for ten min. Finally, 200 μL of the upper-organic phase was collected and 1 μL was directly injected into the GC-MS system. Calibrators and controls were prepared using the same pretreatment procedure. At the analysis of real samples, 100 μL of methanol solvent was added before the extraction solvent.

2.5. Software

Data acquisition and analysis was performed by Agilent Chemstation. Data deconvolution was performed by AMDIS (Automated Mass Spectral Deconvolution and Identification Software), which aided in spectrum “cleaning” by correcting the spectral skew and by determining a more accurate apex retention time (RT). A library search was then performed either by MSD ChemStation or by AMDIS, combining retention time and 4-ion agreement with the processed spectra and, finally, the target peaks were recognized among interferences and reported only if the quality match factor exceeds a preset value.

Retention time lock (RTL) was applied in order to avoid RT shifts due to a variety of reasons such as column trimming, installation of new columns, or other routine procedures. By this approach, RTs were comparable, and the results were more reproducible between the two different laboratories. For RTL, an easily identifiable compound with symmetrical eluting peaks at a crucial point of the chromatogram was chosen and was injected five times with different pressures (±20%, ±10%, and target pressure). The RTL file was created by the Chemstation software and used in every analysis. In the present study, lidocaine was used as the RTL reference compound.

Deconvolution Reporting Software (DRS), which combines AMDIS and MSD ChemStation output, was used to aid the identification of co-eluting compounds and to improve and accelerate the process of the identification of detected peaks.

3. Results

A chromatographic analysis of the 41 analytes was carried out over a 21-min run. All the compounds of interest were separated efficiently in the first 18 min, as depicted in the extracted ion chromatograms presented in Figure 1.

A 10-min backflush program at 300 °C was applied after the thermal gradient provided an adequate purging of the system and aided in the suppression of background noise for increased detection sensitivity.

3.1. Solvent Extraction Selection

The evaluation of the most appropriate extraction solvent, showing efficient extraction recovery for all studied analytes, was performed by studying seven different, commonly used organic solvents. More specifically, butyl acetate, 1-octanol, 4-methyl-2-pentanone, chlorobutane, hexane, diethyl ether, and ethyl acetate were used for analytes extraction. For this study, 1 mL of a blood sample spiked with 0.125 μg/mL of flunitrazepam and fentanyl; 0.25 μg/mL of ropivacaine, bupropion, venlafaxine, mirtazapine, biperiden, citalopram, midazolam, olanzapine, zolpidem, and haloperidol; 1 μg/mL of alprazolam, tramadol, atropine, moclobemide, 7-AF, propofol, ketamine, lidocaine, methadone, amitriptyline, cocaine, chlomipramine, diazepam, nordazepam, and clozapine; and 2 μg/mL of methamphetamine, MDMA, MDEA, MBDB, nortriptyline, codeine, chloropromazine, fluoxetine, sertraline, quetiapine, amantadine, MDA, phenobarbital, and phenytoin were used. The sample was treated as described in the experimental section with the different solvents and the recoveries were assessed.

Based on the experimental observation, ethyl acetate was not considered appropriate and was the first to be rejected due to the notable formation of emulsion, which did not enable the collection of a clear organic phase. Diethyl ether, hexane, and 1-octanol also formed emulsions, however to a lesser extent, thus they were further considered. Besides, it was observed that diethyl ether, due to its low boiling point, evaporated to some extent during the procedure and this is a fact that should be taken into account, as it can introduce bias into the method.

Regarding the recoveries obtained by each solvent, these are depicted in Figure 2 in a comparative way through a heat map. The darker color corresponds to the increased recovery values of the analytes. It can be understood that based on the various chemical structures and properties of the studied analytes, the solvent providing the highest recovery differs for many analytes. For this, the solvent that provided increased recovery for the majority of the analytes was selected as the optimum. Such a compromise was necessary in order to be able to simultaneously extract all analytes efficiently in a single step. It was found that butyl acetate extracted 31 out of 41 analytes with a recovery R% > 85%.

In an attempt to obtain more meaningful conclusions based on the extraction recoveries, a trend was observed according to the separate drug categories. For this, the analytes were grouped into categories. It can be seen that amphetamines were better extracted with chlorobutane, opiates, antidepressants, and anesthetics by butyl acetate; benzodiazepines by butyl acetate and 4-methyl-2-pentanone; antiparkinsonians by butyl acetate and diethyl ether; and, finally, antipsychotics by butyl acetate, 4-methyl-2-pentanone, and chlorobutane. Overall, it can be observed that butyl acetate gives the most satisfying result in almost all categories.

It was seen that diethyl ether provided the lowest recoveries for almost all the analytes and this could be due to losses during the procedure as evaporation was observed in the compounds. 1-octanol was also found to provide low recovery values for all the compounds, with the exception of MDEA, which was extracted with its highest recovery of 64.2% under these conditions. As it concerns the analytes, methamphetamine was the only case where the recoveries were similar to the solvents. On the other hand, phenobarbital and phenytoin showed very low extraction rates < 28.1%, with every solvent tested because of their extremely high pKa values, which are 7.3 and 8.3, respectively. It was also observed that small-molecular-weight drug analytes such as propofol and bupropion were more efficiently extracted with hexane and iso-octane, although the recoveries of the majority of all the other drugs were not favored with these two solvents.

By comparing recovery rates for the whole set of analytes, it can be concluded that 4-methyl-2-pentanone provides a similar extraction efficiency to butyl acetate (25 out of 41 analytes, R% > 85%). However, it was observed that 4-methyl-2-pentanone has a rich mass spectrum in the m/z 58 fragment. This fragment is very common in many of the target analytes and there is a great possibility to produce analytical and chromatographical difficulties, and mainly in cases where the compounds of interest are in low concentrations. More specifically, it effects a high background noise, thus m/z 58 and 100 are basic ions in the mass spectrum of the solvent. This makes the quantification of a significant number of compounds such as some amphetamines, tramadol, and others, which have the ion 58 m/z in abundance and are used for quantitation, more difficult.

Regarding the chromatographic baseline, the solvents 1-octanol, 1-chlorobutane, and 4-methyl-2-pentanone have a high baseline in the first 9 min of analysis. Among the three solvents, 1-octanol produces the highest baseline because it elutes slowly from the column due to its higher molecular weight compared to the others.

Taking into account all of the above-mentioned findings, it can be concluded that butyl acetate is the most appropriate out of the seven tested extraction solvents for this specific panel of analytes, as the majority of them were extracted satisfactorily with R% > 85%. Thus, further optimization was applied for this.

3.2. Sample-to-Solvent Ratio Optimization

In a second step, the solvent-to-blood-sample volume ratio was studied. In general, for all four tested ratios the recoveries of the analytes were satisfactory, and in all cases the majority was > 85%. However, a sample-to solvent-ratio of 4:1 (v/v) and 1:1 (v/v) provided the highest recoveries for the majority of the analytes. In Figure 3, the recoveries R% for all analytes under the four extraction conditions are presented comparatively in a heat map. For some cases of analytes, such as amantadine, atropine, codeine, and alprazolam, their recovery increased as the solvent-to-blood-volume ratio decreased. Under 4:1 and 1:1 ratios, all categories of drugs, e.g., benzodiazepines, antidepressants, etc., were efficiently extracted. Among these, the 4:1 (v/v) was selected as it has the most satisfying recovery results for almost all the compounds when using a reasonable volume of blood sample. In Table S1, all the volumes used in the sample-to-volume ratio experiment are presented.

3.3. Method Validation

3.3.1. Selectivity

Blood samples from six different drug-free subjects were tested for the presence of endogenous components, which might interfere with the detection of the 41 drugs and DOA or the internal standard. It was found that no endogenous blood components were eluted at the same retention times with the analytes.

3.3.2. Carryover

A blank sample was analysed after six continuous analyses of spiked samples at the highest concentration, and it was proven that no carryover effect occurred for any of the analytes.

3.3.3. Linearity

The linearity of the method was studied within the range of the mean therapeutic and the mean toxic concentrations of each drug, which were found in the literature [40]. Drug-free blood samples were spiked at five concentrations levels with 100 μL of methanolic solutions and 25 μL of IS, then analysed. Every standard was analysed in four replicates and the calibration curves were constructed based on the peak areas ratio of the analyte to the IS, which were linear in the studied ranges with a correlation coefficient that was higher than 0.9934 for all analytes (see

Table 2. Figures of merit of the method for 41 studied analytes.

Analyte	Linear Range (μg/mL)	Linear Equation	R2	LOD (μg/mL)	LOQ (μg/mL)
Methampetamine	0.4–10.0	y = 10.101x + 1.4706	0.9978	0.020	0.066
Amantadine	0.8–20.0	y = 4.3684x + 1.559	0.9994	0.010	0.032
Propofol	0.1–5.0	y = 24.617x − 2.6623	0.9983	0.005	0.017
MDA	0.8–20.0	y = 1.6608x − 0.0176	0.9990	0.033	0.109
MDMA	0.4–10.0	y = 11.137x + 0.7223	1.000	0.009	0.031
MDEA	0.4–10.0	y = 20.563x + 1.4997	0.9999	0.002	0.008
Bupropion	0.05–1.00	y = 17.6133 + 0.2203	0.9991	0.013	0.044
MBDB	0.4–10.0	y = 24.32x + 0.8195	0.9997	0.002	0.007
Fluoxetine	0.5–10.0	y = 26.587x − 16.212	0.9968	0.055	0.184
Ketamine	0.1–5.0	y = 2.9372x + 0.3221	0.9957	0.011	0.038
Lidocaine	0.1–5.0	y = 5.7228x + 0.7541	0.9955	0.009	0.029
Tramadol	0.2–5.0	y = 33.57x − 1.1145	0.9995	0.001	0.004
Phenobarbital	0.4–20.0	y = 12.865x + 5.734	0.9987	0.008	0.027
Venlafaxine	0.02–1.00	y = 44.847x − 0.4172	0.9995	0.003	0.011
Methadone	0.1–5.0	y = 37.989x + 0.4546	0.9997	0.003	0.011
Ropivacaine	0.08–2.00	y = 32.713x − 0.4533	0.9998	0.003	0.009
Amitriptyline	0.1–5.0	y = 35.265x − 0.4943	0.9978	0.012	0.040
Cocaine	0.1–5.0	y = 11.402x − 0.6845	0.9976	0.006	0.019
Atropine	0.2–5.0	y = 3.1589x + 0.2277	0.9934	0.011	0.037
Nortriptyline	0.75–10.00	y = 12.427x − 8.7535	0.9996	0.113	0.375
Moclobemide	0.2–5.0	y = 25.376x + 1.6707	0.9996	0.003	0.009
Mirtazapine	0.02–1.00	y = 28.645x − 0.0298	0.9988	0.006	0.020
Biperiden	0.02–1.00	y = 34.649x − 0.5564	0.9995	0.003	0.009
Phenytoin	0.4–20.0	y = 9.9039x − 4.2309	0.9979	0.012	0.041
Sertraline	0.5–10.0	y = 4.3317x − 1.7901	0.9989	0.075	0.251
Citalopram	0.05–1.00	y = 24.7065 − 0.5613	0.9989	0.003	0.010
Codeine	0.4–10.0	y = 4.5406x + 0.8487	0.9997	0.002	0.007
Clomipramine	0.1–5.0	y = 12.546x + 1.8696	0.9981	0.007	0.024
Diazepam	0.1–5.0	y = 8.8033x − 0.5545	0.9987	0.014	0.045
Chlorpomazine	0.4–10.0	y = 16.567x + 1.9159	0.9998	0.003	0.011
Nordazepam	0.1–5.0	y = 10.1663x + 0.6575	0.9986	0.012	0.040
Midazolam	0.02–1.00	y = 30.607x − 0.3161	0.9975	0.002	0.007
Flunitrazepam	0.025–1.000	y = 2.9896x − 0.0529	0.9999	0.003	0.011
7-AF	0.35–5.00	y = 1.5609x − 0.0223	0.9993	0.037	0.128
Fentanyl	0.01–0.50	y = 14.9604x − 0.1360	0.9977	0.003	0.010
Olanzapine	0.02–1.00	y = 6.4316x − 0.2464	0.9974	0.004	0.011
Zolpidem	0.02–1.00	y = 20.3231x + 0.3246	0.9975	0.003	0.011
Clozapine	0.1–10.0	y = 6.9347x − 0.4942	0.9992	0.011	0.037
Haloperidol	0.1–5.0	y = 3.1195x − 0.6406	0.9961	0.028	0.093
Alprazolam	0.05–1.00	y = 4.1498x − 0.1051	0.9986	0.009	0.026
Quetiapine	0.5–10.0	y = 1.4263x − 0.5078	0.9989	0.054	0.174

).

3.3.4. Limits of Detection and Quantification

The LOD and LOQ were experimentally calculated as a signal-to-noise ratio of 3:1 and 10:1, respectively, for every drug, and were found to range between 1–113 and 4–375 ng/mL, respectively—as can be seen in Table 2. The reported values for all pharmaceuticals were lower or close to the lower therapeutic values, whereas for DOA they were quite low.

3.3.5. Accuracy and Precision

The results for accuracy and precision within a batch (intraday) and over a period of a week (interday) at three quality control levels (LQC, MQC, HQC), which represented the entire corresponding dynamic range of the calibration curve for each analyte, are presented in

Table 3. Intra- and interassay of the compounds.

Compound	Added (μg/mL)	Intra-Assay				Inter-Assay
		Mean Found (μg/mL)	SD	CV %	Accuracy %	Overall Mean Found (μg/mL)	SD	CV %	Accuracy %
Methamphetamine	0.50	0.51	0.03	8.03	103	0.51	0.02	5.84	103
	2.50	2.47	0.03	1.21	99	2.50	0.04	1.76	100
	8.00	8.02	0.21	2.09	100	7.84	0.32	3.28	98
Amantadine	0.90	0.90	0.06	7.36	100	0.89	0.05	6.32	99
	5.00	4.94	0.20	4.13	99	4.94	0.17	3.44	99
	18.0	18.0	0.63	3.17	100	17.4	1.06	5.47	97
Propofol	0.40	0.44	0.01	5.10	110	0.45	0.01	5.02	112
	1.25	1.17	0.08	6.84	94	1.19	0.12	10.4	95
	4.00	4.16	0.40	7.60	104	4.20	0.49	9.31	105
MDA	0.90	0.94	0.07	8.12	105	0.91	0.06	6.90	102
	5.00	5.25	0.13	2.40	105	5.00	0.35	6.94	100
	18.0	18.1	0.52	2.57	100	18.0	0.45	2.27	100
MDMA	0.50	0.50	0.03	6.50	100	0.50	0.02	4.71	101
	2.50	2.53	0.05	1.78	101	2.52	0.05	1.90	101
	8.00	8.01	0.22	2.20	100	7.98	0.38	3.85	100
MDEA	0.50	0.52	0.03	6.69	105	0.52	0.02	5.04	104
	2.50	2.57	0.10	3.89	103	2.57	0.08	2.96	103
	8.00	7.96	0.29	2.92	100	7.92	0.31	3.10	99
Bupropion	0.06	0.07	0.01	9.09	110	0.06	0.01	14.8	100
	0.25	0.26	0.04	14.5	105	0.25	0.03	11.4	98
	0.80	0.75	0.08	7.95	94	0.77	0.10	10.4	96
MBDB	0.50	0.52	0.02	4.61	103	0.52	0.02	3.86	104
	2.50	2.58	0.06	2.28	103	2.56	0.05	1.99	102
	8.00	7.92	0.24	2.42	99	7.96	0.21	2.06	99
Fluoxetine	1.30	1.33	0.09	7.19	102	1.31	0.10	7.70	101
	5.00	5.09	0.51	9.98	102	4.89	0.43	8.81	98
	8.00	8.65	1.05	9.69	108	8.26	0.96	9.26	103
Ketamine	0.30	0.28	0.02	9.13	92	0.27	0.03	12.5	90
	1.25	0.21	0.09	7.86	90	1.07	0.08	7.68	85
	4.00	4.08	0.31	6.00	102	3.93	0.42	8.62	98
Lidocaine	0.30	0.30	0.03	9.92	101	0.28	0.03	11.8	95
	1.25	1.31	0.09	7.02	105	1.35	0.12	8.70	108
	4.00	3.97	0.27	5.40	99	4.06	0.51	9.98	101
Tramadol	0.30	0.33	0.01	5.07	109	0.33	0.01	5.43	111
	1.25	1.29	0.02	1.78	104	1.28	0.03	2.34	103
	4.00	3.95	0.07	1.40	99	3.96	0.10	2.10	99
Phenobarbital	1.25	1.14	0.12	12.7	91	1.19	0.12	13.1	95
	5.00	5.48	0.31	5.59	110	5.28	0.35	6.55	106
	19.0	19.9	1.67	8.00	105	19.7	1.78	8.60	104
Venlafaxine	0.06	0.07	0.01	12.7	110	0.06	0.01	13.2	106
	0.25	0.24	0.02	9.92	97	0.24	0.02	8.33	96
	0.80	0.77	0.07	7.00	96	0.81	0.10	10.3	101
Methadone	0.30	0.28	0.02	8.16	93	0.27	0.02	10.2	90
	1.25	1.28	0.08	6.18	102	1.28	0.07	5.15	103
	4.00	4.14	0.40	7.75	103	4.08	0.42	8.23	102
Ropivacaine	0.09	0.10	0.003	3.33	113	0.10	0.003	3.37	111
	0.50	0.50	0.02	4.04	99	0.50	0.01	2.63	99
	1.90	1.89	0.04	2.11	99	1.88	0.04	1.87	99
Amitriptyline	0.30	0.33	0.02	5.90	108	0.31	0.03	10.3	105
	1.25	1.29	0.14	10.9	103	1.25	0.16	12.4	100
	4.00	4.07	0.73	14.3	102	3.97	0.51	10.2	99
Cocaine	0.30	0.33	0.01	2.18	110	0.33	0.03	9.49	110
	1.25	1.13	0.07	6.31	90	1.12	0.09	7.86	90
	4.00	4.25	0.65	12.2	106	4.17	0.57	10.9	104
Atropine	0.30	0.30	0.01	5.45	101	0.30	0.01	3.96	101
	1.25	1.42	0.02	1.20	114	1.37	0.07	5.25	110
	4.00	3.94	0.14	2.87	98	3.95	0.11	2.31	99
Nortriptyline	0.80	0.81	0.01	0.66	101	0.80	0.01	0.80	100
	2.50	2.42	0.08	3.30	97	2.40	0.09	3.71	96
	9.00	9.85	0.75	6.85	109	9.98	0.63	5.67	111
Moclobemide	0.30	0.30	0.01	4.46	101	0.30	0.01	3.96	101
	1.25	1.29	0.09	6.69	103	1.29	0.07	5.34	103
	4.00	3.97	0.08	1.69	99	4.00	0.10	2.00	100
Mirtazapine	0.06	0.07	0.007	12.07	116	0.06	0.007	14.0	100
	0.25	0.26	0.025	9.80	102	0.26	0.034	13.2	103
	0.80	0.73	0.096	10.51	91	0.79	0.105	10.6	99
Biperiden	0.06	0.06	0.003	6.25	96	0.06	0.004	8.00	100
	0.25	0.25	0.008	3.20	100	0.25	0.007	2.83	99
	0.80	0.75	0.065	6.94	94	0.77	0.053	5.51	96
Phenytoin	1.25	1.21	0.117	12.10	97	1.19	0.115	12.0	96
	5.00	5.44	0.222	4.08	109	5.49	0.437	7.97	110
	19.0	19.2	2.114	10.45	101	19.5	2.113	10.3	103
Sertraline	1.30	1.01	0.073	6.92	84	1.13	0.093	8.56	87
	5.00	5.51	0.490	8.90	110	5.12	0.544	10.6	102
	8.00	8.61	1.119	10.40	108	8.27	0.955	9.24	103
Citalopram	0.06	0.06	0.005	10.47	96	0.07	0.007	13.0	108
	0.25	0.26	0.036	13.85	104	0.23	0.037	14.5	102
	0.80	0.85	0.091	8.58	106	0.81	0.091	8.96	102
Codeine	0.50	0.49	0.019	4.82	99	0.50	0.021	5.29	99
	2.50	2.57	0.035	1.36	103	2.57	0.057	2.22	103
	8.00	7.86	0.162	1.65	98	7.92	0.221	2.23	99
Clomipramine	0.30	0.28	0.022	9.61	92	0.27	0.019	8.37	91
	1.25	1.16	0.069	5.93	93	1.20	0.072	5.99	96
	4.00	3.75	0.269	5.74	94	3.85	0.289	6.00	96
Diazepam	0.30	0.33	0.042	15.50	108	0.31	0.039	15.2	103
	1.25	1.31	0.061	4.66	105	1.25	0.103	8.26	100
	4.00	4.20	0.402	7.60	105	4.06	0.466	9.19	101
Chlorpromazine	0.50	0.50	0.017	4.25	100	0.51	0.016	3.92	102
	2.50	2.53	0.042	1.66	101	2.51	0.040	1.59	100
	8.00	7.98	0.165	1.66	100	8.03	0.184	1.83	100
Nordazepam	0.30	0.31	0.04	12.86	104	0.31	0.04	12.9	103
	1.25	1.24	0.08	6.52	99	1.25	0.09	7.22	100
	4.00	4.11	0.26	6.32	103	4.08	0.31	7.60	102
Midazolam	0.06	0.07	0.003	5.36	112	0.07	0.005	8.50	118
	0.25	0.21	0.010	4.76	84	0.22	0.020	8.93	88
	0.80	0.74	0.070	7.53	93	0.80	0.100	10.0	100
Flunitrazepam	0.025	0.03	0.003	11.11	108	0.03	0.002	7.41	108
	0.125	0.12	0.014	12.17	92	0.12	0.011	9.48	93
	0.40	0.40	0.030	6.04	99	0.40	0.022	4.41	100
7-AF	0.40	0.40	0.032	9.20	99	0.40	0.023	6.63	99
	1.25	1.20	0.037	3.09	96	1.21	0.036	2.98	97
	4.00	4.01	0.085	1.69	100	4.01	0.074	1.48	100
Fentanyl	0.03	0.03	0.003	11.54	104	0.03	0.003	11.5	104
	0.125	0.11	0.006	5.61	86	0.11	0.008	7.41	86
	0.40	0.36	0.046	10.22	90	0.39	0.050	10.4	96
Olanzapine	0.06	0.07	0.007	12.07	116	0.07	0.005	8.77	114
	0.25	0.26	0.025	9.80	102	0.23	0.023	10.2	90
	0.80	0.73	0.096	10.51	91	0.79	0.091	9.27	98
Zolpidem	0.06	0.06	0.007	13.73	102	0.06	0.007	14.6	96
	0.25	0.22	0.021	9.68	87	0.22	0.017	7.83	87
	0.80	0.83	0.078	7.57	103	0.79	0.086	8.73	99
Clozapine	0.30	0.26	0.014	6.51	86	0.28	0.023	9.75	94
	1.25	1.26	0.046	3.66	101	1.24	0.068	5.49	99
	4.00	4.02	0.297	6.18	100	4.02	0.475	9.46	100
Haloperidol	0.30	0.33	0.016	5.78	111	0.33	0.016	5.80	110
	1.00	1.09	0.110	10.13	87	1.07	0.082	7.66	86
	4.00	3.82	0.369	7.72	96	3.94	0.516	10.5	98
Alprazolam	0.06	0.06	0.005	10.42	96	0.06	0.009	18.0	100
	0.25	0.21	0.010	4.69	85	0.22	0.020	8.93	90
	0.80	0.69	0.070	8.12	86	0.79	0.110	11.2	98
Quetiapine	1.30	1.21	0.068	5.84	93	1.23	0.071	5.98	95
	5.008.000	4.90	0.108	2.20	98	5.06	0.367	7.25	101
	8.00	8.09	0.618	6.11	101	7.90	0.738	7.48	99

. The accuracy was found to be between 85% and 114% and the precision was less than 14.8%.

The interlaboratory precision of the method was also examined by the analysis of twelve real samples, which were found to be positive in more than one drug in our laboratory (Laboratory of Forensic Toxicology of the University of Thessaloniki) and in the Laboratory of Forensic Service of Ministry of Justice. Seventeen different compounds were detected and determined by both labs in all samples. The concentration results were found to be in good agreement; the correlation coefficient was calculated and found to be 0.9993. The results can be seen in

Table 4. Interlaboratory results.

Sample (S)	Compounds	Concentration Found in Laboratory A (μg/mL)	Concentration Found in Laboratory B (μg/mL)
S1	Tramadol	0.271	0.270
	Codeine	0.018	0.018
	Flunitrazepam	0.289	0.236
	7-Aminoflunitrazepam	0.707	0.730
	Alprazolam	0.876	0.781
	Lidocaine	0.171	0.174
S2	Zolpidem	5.00	4.50
	Tramadol	0.253	0.230
S3	Ropivacaine	2.00	2.00
	Lidocaine	0.988	0.955
S4	Quetiapine	15.7	15.2
	Diazepam	0.028	0.025
S5	Diazepam	0.283	0.273
	Propofol	0.901	0.738
	Lidocaine	0.049	0.043
	Midazolam	0.086	0.083
S6	Sertraline	0.308	0.270
S7	Clozapine	2.50	2.30
	Lidocaine	0.051	0.051
S8	Clozapine	4.200	4.100
	Lidocaine	78	74
S9	Clozapine	4.20	4.40
	Lidocaine	0.066	0.061
S10	Mirtazapine	0.050	0.045
	Citalopram	0.206	0.192
S11	Clozapine	4.80	4.90
	Zolpidem	0.062	0.059
	Diazepam	0.580	0.559
	Biperiden	0.142	0.141
S12	Diazepam	0.038	0.039
	Propofol	1.50	1.60
	Midazolam	0.058	0.055
	Lidocaine	0.034	0.030

.

4. Results and Discussion of Real Sample Analysis

The method has been applied for hundreds of samples for the investigation of postmortem and clinical cases. It has been proven to be a very valuable, rapid, and low-cost tool in the routine analysis of our laboratory for the monitoring of various drugs and DOA. Hundreds of samples have been analysed, however, for the economy of the presentation, a number of these cases have been chosen. Those with a history of pharmaceutical use for mental disorders from January to March 2021 have been selected to be presented here for exemplification purposes and to provide proof of concept.

Postmortem blood samples from 89 cases of forensic interest (history of mental disorders and use of psychiatric pharmaceuticals) were analyzed with the developed method and provided valuable data.

As an aggregate, the number of positive detections of the monitored drugs and pharmaceuticals in the 89 samples reached a total sum of 135. Eight pharmaceuticals out of the forty-one under study were found and determined in the examined samples. These comprised drugs belonging to different categories (benzodiazepines, antidepressants, antipsychotics, etc.). In

Table 5. Concentration in μg/mL of the determined pharmaceutical in each case.

Case No	Diazepam	Citalopram	Alprazolam	Olanzapine	Mirtazapine	Venlafaxine	Haloperidol	Zolpidem
1	0.34					0.43
2	1.12
3	0.62	0.17
4	1.30
5	0.52					0.18
6	0.18
7	0.76		0.02	0.013
8	0.13
9	0.74
10	0.25
11	0.14
12	0.26				0.07
13	0.93
14	0.79
15	0.51				0.12
16	0.35
17	1.01
18	1.16
19	4.34					0.24
20	0.66
21	0.59		0.12
22	0.67
23	0.24
24	0.76				0.16
25	0.56					0.33
26	0.19	0.2
27	0.63			0.024
28	0.21					0.006
29	0.16
30	0.18				0.28
31	0.25	0.04
32	3.50
33	0.69					0.03
34	0.52
35	2.88
36	0.21
37	0.24
38	0.26
39	0.22			0.029
40	0.14	0.19			0.33
41	0.23	0.05					0.13
42	0.20
43	0.14
44	0.27	0.11		0.009
45	0.22
46	0.27	0.12
47	0.41	0.06		0.083
48	0.27	0.21	0.05		0.06
49	0.60
50	0.32	0.17
51	0.22			0.032
52	0.13
53	0.43	0.13	0.03
54	0.31		0.02
55	0.15						0.05
56	0.26	0.05
57	0.32			0.015
58		0.04
59		0.16	0.02
60		0.05
61		0.24	0.08
62		0.15
63		0.14
64			0.05
65			0.01		0.05
66			0.01
67			0.09		0.03
68			0.06
69		0.12		0.035
70				0.028
71				0.010
72				0.022
73				0.017
74					0.05			0.01
75					0.10
76					0.24			0.16
77					0.03
78					0.01
79						0.63	0.015
80						0.92
81						0.01
82						0.22
83						0.06
84						0.15
85						0.19
86							0.007
87								0.10
88								0.09
89								0.11

, the concentrations of the pharmaceuticals in every case are given, showing the potential and effectiveness of the method. The frequency in which they were found is outlined schematically in Figure 4. The majority of the identified compounds were benzodiazepines (74 out of 135), which is in agreement with the metadata, considering that the samples were obtained by individuals with mental illnesses.

By comparing the concentration values found with those referred to in the international literature [40], these generally appeared to be at therapeutic levels. More specifically, the concentration ranges of the compounds found in the studied cases ranged for diazepam from 0.13 to 4.34 μg/mL, for citalopram from 0.04 to 0.24 μg/mL, for alprazolam from 0.01 to 0.12 μg/mL, for olanzapine from 0.009 to 0.083 μg/mL, for mirtazapine from 0.01 to 0.33 μg/mL, for venlafaxine from 0.006 to 0.92 μg/mL, for haloperidol from 0.007 to 0.13 μg/mL, and for zolpidem from 0.01 to 0.16 μg/mL. In addition, all these positive results in low concentrations have been gained with a single LLE step, which is in contrast with other analytical methods that had similar findings with much more time-consuming steps, such as SPE [25,27].

The use of sophisticated embedded software such as DRS, where a report of candidate compounds that combines deconvoluted peaks and retention time data from RTL, is also an important feature of the methodology. However, analysts should be extremely cautious in the interpretation of the results; when the concentration of the compound is near the LOD of the method, the generated report may not include the specific compound, thereby leading to false negative results. Thus, aid by the software should be wisely used and the examination of the raw data is very important to confirm findings.

5. Conclusions

A rapid, sensitive, and reliable method for the simultaneous detection of 41 drugs and DOA in blood was developed. The method provides detection and accurate quantitation whenever needed. Fast methods that apply LLE without evaporation or derivatization steps are very limited and such a methodology is of high interest for laboratories performing routine drug analysis. Also, mostly all methods used for screening are not validated, thus Limits of Detection of the analytes are a blur or not strictly determined. The reported method was validated and figures of merit were determined. A paradigm of application is presented by providing data for a number of cases showing the utility of the method.