Talanta 241 (2022) 123265

Contents lists available at ScienceDirect

Talanta
journal homepage: www.elsevier.com/locate/talanta

Development and validation of a UHPLC-HRMS-QTOF method for the

detection of 132 New Psychoactive Substances and synthetic opioids,
including fentanyl, in Dried Blood Spots
Marta Massano a, *, Carola Incardona a, Enrico Gerace b, Pierre Negri c, Eugenio Alladio a,
Alberto Salomone a, b, Marco Vincenti a, b
a
Department of Chemistry, University of Torino, Via P. Giuria 7, 10125, Torino, Italy
b
Centro Regionale Antidoping e di Tossicologia “A. Bertinaria”, Regione Gonzole 10/1, 10043, Orbassano, Torino, Italy
c
SCIEX, Redwood City, CA, 01701, USA

A R T I C L E I N F O A B S T R A C T

Keywords: Dried Blood Spots (DBS) represents a promising micro-sampling technique in the field of forensic toxicology to
NPS carry out minimally invasive blood sample collection. In DBS, cheap, fast and easy sampling is combined with
Fentanyl effortless store and transport. These properties aimed us to develop and validate a quick and easy procedure for
Dried blood spots
the detection of a large and diverse range of emerging and alarming New Psychoactive Substances (NPS). A drop
Validation
of whole blood sample was collected on a DBS card and dried for 3 h, from which a total of 132 analytes
QTOF
HRMS (including NPS, synthetic opioids NSO and metabolites) plus 13 deuterated internal standards could be extracted
Opioids using 500 μL of a methanol/acetonitrile mixture (3:1, v/v) and subsequently separated and identified by means
of ultra-high-performance liquid-chromatography (UHPLC) coupled to high resolution mass spectrometry
(HRMS). The extraction efficiency proved to be reproducible with yields ranging from 30% to 100% depending
on the different classes of drugs. Trueness, repeatability, and intermediate precision fulfilled acceptance criteria
for almost all synthetic opioids, cathinones and hallucinogens (bias and CV% below ±20%); in particular, the
aggregate inter-day trueness data showed extremely limited deviation from the expected concentrations (− 10%
< bias% < +10%) for 114 target analytes out of 132. The calculated limits of detection ranged from 1.3 to 6.3
ng/mL, consistently exceeding the values experimentally tested. Moderate ion suppression was observed for most
analytes, partly caused by blood fortification itself. Good stability of the target analytes at − 20 ◦ C, 4 ◦ C, and
35 ◦ C on DBS cards after drying was observed, even for long periods of time. Optimal storage condition appeared
to be at 4 ◦ C resulting in virtually no drugs degradation for up to 40 days. The novel analytical method based on
DBS sampling, verified on venous whole blood real samples previously tested positive with our routine pro­
cedure, conveys remarkable potential in analytical toxicology, clinical analysis, and doping control.

1. Introduction synthetic opioids, cathinones and hallucinogen, but this simplistic

classification does not express adequately the variety and complexity of
Drug abuse has certainly represented a dramatic social and health their potency, combined effects, and risk profiles that intersect cate­
issue for a long time, but in recent years the diffusion of countless New gories and often differentiate compounds belonging to the same cate­
Psychoactive Substances (NPS) in the illegal drug market introduced an gory [4].
unprecedented threat to the public health [1]. The adjective “new” does Nowadays, the NPS abuse in U.S and Europe is proliferating at un­
not necessarily indicate totally original compounds, but it may refer to precedented rate [5] and represents an increasing challenge to the
substances initially synthesized and tested for their potential pharma­ established national and international drug policies. Difficulties associ­
ceutical properties and revalued years later as highly potent drugs of ated with NPS monitoring include their high number, the speed with
abuse [2,3]. NPS are often categorized into synthetic cannabinoids, which they enter and exit the illegal drug market, the restricted and

* Corresponding author. Department of Chemistry, University of Turin, via Pietro Giuria 7, 10125, Turin, Italy.
E-mail address: [email protected] (M. Massano).

https://doi.org/10.1016/j.talanta.2022.123265
Received 3 December 2021; Received in revised form 21 January 2022; Accepted 24 January 2022
Available online 29 January 2022
0039-9140/© 2022 Elsevier B.V. All rights reserved.
M. Massano et al. Talanta 241 (2022) 123265

partial information about them, the incomplete knowledge about their The blank samples were then fortified at six concentration levels (5, 7.5,
composition and mixing, the variable and often unknown potency, and 10, 25, 50, 100 ng/mL) with an NPS working solution.
the difficulty for toxicological laboratories to find the analytical stan­ The blood samples were vigorously stirred prior to spotting a 30 μL
dards to provide confirmatory testing [6]. New analytical approaches aliquot on a DBS collection card using a calibrated pipette. The spots
and in-depth investigations of the different biological matrices, either were allowed to dry for at least 3 h at room temperature. The entire spots
conventional or alternative, are thus needed for effective monitoring, on the DBS card were punched out and transferred into an Eppendorf
surveillance, drug control, and public health campaigns aimed to reduce tube together with 500 μL methanol/acetonitrile (3:1, v/v) mixture. A
this drug-related harm [7]. 12.5 μL solution of the deuterated internal standards (ISTD) was added
Dried blood spot (DBS) is a sampling technique based on the to the tubes to reach a final concentration of 25 ng/mL. After ultra­
collection of a whole blood drop on a filter paper. It has been widely sonication (215/860 W, 50/60 Hz) for 30 min at room temperature, the
used in the diagnostic screening of neonatal metabolic disorders [8] and tubes were centrifuged for 5 min at 13,000 g. The supernatant was
is now available for clinical diagnostics [9] as well as other applications, transferred into a fresh tube prior to evaporating the solvent under a
including forensic toxicology [10,11] and antidoping analyses [12]. gentle nitrogen flow at room temperature. The dry residue was recon­
With the development of modern mass spectrometers, several studies stituted with 50 μL methanol, centrifuged for 5 min at 13,000 g, and 5 μL
have been published on the use of DBS to monitor the blood concen­ of the supernatant was injected into the UHPLC system.
tration of a single or multiple target substances for neonatal diagnosis
[13], drugs of abuse [14] and therapeutic drugs [15]. The collection of 4. Instrumentation
whole blood dried on a piece of filter paper provides several advantages
including simplicity, speed, resistance to manipulations, and enhanced UHPLC separation was performed with a Phenomenex Kinetex C18
stability of the target analytes at room temperature. Compared to con­ column (100 × 2.1 mm, 1.7 μm) maintained at 45 ◦ C on the SCIEX
ventional venous blood sampling, the typical 30–50 μL collection of ExionLC™ AC system. The mobile phases consisted of water (A) and
capillary blood, obtained by finger, ear, or heel prick, is minimally acetonitrile (B), both with formic acid 5 mM. The LC flow rate was set at
invasive. The minimum space occupied by the sampling paper as well as 0.5 mL/min and the mobile phase eluted under the following linear
the reduced effects of environmental degradation result in a facilitated gradient conditions: (A:B, v:v) isocratic elution at 95:5 for 0.5 min, from
storage and shipment of DBS specimens [16]. The continuous sensitivity 95:5 to 5:95 in 7.5 min, isocratic elution at 5:95 for 0.5 min and final re-
improvement enabled by mass spectrometric techniques encourages equilibration for 2.5 min to the initial condition. The total run time was
further a minimal blood volume handling, such as the one involved in 10 min.
DBS [17]. All analyses were performed using a quadrupole/time-of-flight
The present study aims to highlight the opportunities and potential SCIEX X500R QTOF mass spectrometer (Sciex, Darmstadt, Germany)
benefits arising from the implementation of DBS as a complementary equipped with a Turbo VTM ion source operating in electrospray
tool in forensic control programs to monitor the NPS diffusion. To this positive-ion mode (full MS and MS/MS parameters available in the
purpose, the present approach combines DBS sampling with ultra-high- Supplementary Table S1). Data acquisition involved a preliminary TOF-
pressure liquid-chromatography (UHPLC) coupled to time-of-flight high MS high-resolution full scan followed by a SWATH™ acquisition pro­
resolution mass spectrometry (TOF-HRMS). The analytical method tocol which used a variable window setup (18 windows covering mass
developed in this study achieved the simultaneous qualitative and range from m/z 99.5 to 575.0 at 0.025 resolving power), resulting in a
quantitative analysis of 132 NPS/NSO analytes and metabolites (their final cycle time of 0.933 s. The qualitative identification of the 132
full list is available in Table 1), with the goal of reaching adequately low target analytes was based on the coincidence of their retention times,
LOQ levels using a simple and rapid extraction procedure. precursor ion and characteristic fragment ion m/z values (mass error
accepted <5 ng/mL) (Table S1) as identification parameters. Data were
2. Reagents and standards acquired using the SCIEX OS 1.5 Software.

All chemicals, including methanol, formic acid and acetonitrile, were 5. Method validation
purchased from Sigma-Aldrich (Milan, Italy). Ultra-pure water was ob­
tained using a Milli-Q® UF-Plus apparatus (Millipore, Bedford, MA, The validation strategy was based on a protocol recently published
USA). All stock standard solutions were prepared in methanol at 1 mg/ [18]. Each standard of blank whole blood fortified at six concentration
mL and stored at − 20 ◦ C until used. Working solution of 132 analytes levels (5, 7.5, 10, 25, 50, 100 ng/mL) was analyzed nine times in three
(identified among the most common synthetic cannabinoids, synthetic working sessions (i.e., 3 × 3) along ten days. This dataset of 54 analysis
opioids, synthetic cathinones and hallucinogens monitored nationally formed the groundwork on which the statistical evaluation of several
and internationally) and internal standard solution containing 13 validation parameters was founded, including calibration, intra- and
deuterated NPSs, were prepared at the final concentration of 1 μg/mL by inter-day trueness, repeatability and intermediate precision (at 6 con­
dilution with methanol. The analytical standards of the target analytes centration levels), limit of detection (LOD), limit of quantification
and deuterated internal standards were purchased from LGC Pro­ (LOQ). Recovery, matrix effect and stability parameters were evaluated
mochem (Milan, Italy) and Sigma-Aldrich (Milan, Italy) (purity >99%, with further independent experiments. Other independent analyses on
concentration between 0.1 mg/mL and 1 mg/mL), or kindly provided by purposely spiked samples were performed later to verify the actual
the National Early Warning System (provided at a concentration of 0.02 detection of the target analytes at the calculated LODs (see the following
mg/mL). DBS cards (FTA™ DMPK C) were purchased from Whatman™ paragraphs). An ad hoc Excel® sheet was built in-house to adapt the
GE Healthcare (UK). routine developed by Desharnais et al. [19]. All the equations employed
to compute the validation parameters have been omitted from this text
3. Sample preparation and can be found elsewhere [20].

For the preparation of spiked specimens used in the analytical 5.1. Calibration
method development and validation, a standard blood matrix was ob­
tained by mixing different aliquots of blank whole blood obtained from Calibration curves were generated from the peak-area ratio between
volunteers (laboratory personnel) after having signed an informed each analyte and the ISTD quantifier transitions; the ratio was then
consent. The absence of all the investigated analytes was verified by plotted on the y-axis against the nominal analyte concentration to
means of standard procedures routinely used in our lab for blood testing. generate and estimate the curve that best fits and predicts the data

2
M. Massano et al. Talanta 241 (2022) 123265

distribution, with the support of statistical tests [21]. In the first step of was expressed as the mean percentage ratio between the two signals
the calibration process, homoscedastic vs. heteroscedastic distribution with its uncertainty expressed as extraction repeatability (CV%) from
of the data points was evaluated by analyzing the residues distribution the three replications. Also in this case, the recovery was measured again
(9 data-points at 6 concentration levels) and comparing the variances at for the analytes whose value was lower than 50% in order to evaluate a
low, medium and high concentration levels. If the variance increased potential correlation between recovery and the compresence of analytes
with concentration, the system proved heteroscedastic (p < 0.05) and a commonly present/absent in the same real sample.
weighted model was adopted, using either a 1/x weighting factor (when
the variance increased proportionally to the concentration) or a 1/x2 5.5. Stability
weighting (when a quadratic increase of the variance was observed).
Then, the order of the calibration model (linear vs. quadratic) was Stability was evaluated at low (10 ng/mL) and high (100 ng/mL)
selected on the basis of Mandel and lack-of-fit tests, using a significance concentrations levels. Following the deposition of the drop of blood and
level of 95%. after 3 h of drying, the samples were placed at the temperature condi­
tions of − 20 ◦ C, 4 ◦ C and 35 ◦ C and stored before analysis for, respec­
5.2. LOD and LOQ tively, 1, 3, 7 and 40 days. The conditions were chosen to simulate the
possible storage scenarios, from the mildest to the most critical condi­
The Hubaux-Vos method [22] was used for the calculation of the tions, so as to evaluate i) the transfer of the cards from the sampling site
LOD values. The original method is based on the hypothesis that the data (e.g., workplace) to the testing laboratory, ii) a short-term storage of the
distribution is homoscedastic. Since this condition is usually not met, the cards before analysis, and iii) a long-term storage of the cards, in the
weighting factors were included in the Hubaux-Vos calculation of the event that a counter-analysis is requested after a long time interval, for
LOD, as described in the Currie’s method [23]. LOQs were attributed to example when a positive result is challenged by the sample donor. The
the lowest concentration within the respective calibration range stability conditions were observed through trend lines that describe the
yielding trueness values within the accepted limit (typically, bias% < variation of the concentration calculated at the different storage tem­
±20%). peratures and after different storage intervals. Since a 15% deviation
The calculated LOD values were then experimentally tested by from the nominal value is compatible with the experimental uncer­
spiking the blank matrix with the target analytes at the approximate tainty, the presence of degradation effects was positively detected above
LOD concentrations and verifying that the signal-to-noise ratio (S/N) this 15% limit [26].
was higher than 3.
6. Application on real samples
5.3. Trueness and precision
The applicability of the developed procedure on real samples was
The procedure used to calculate the intra-day trueness (expressed as verified on venous whole blood samples previously analyzed and tested
bias %) considers separately the three calibration curves of the same positive by means of a previously published UHPLC-MS/MS method,
day: two of these are cyclically used for the construction of a calibration which is periodically reviewed and updated according to new com­
model, which is used to perform a back-calculation of the points of the pounds entering the market [27]. At the time of analysis, the blood real
third curve. The concentration values calculated by repeating the pro­ specimens were obtained by venous sampling and then stored until
cedure for the three days are used to determine the bias %. The calcu­ analysis under controlled conditions, at the temperature of 4 ◦ C. A drop
lation of the inter-day trueness is similar, but the calibration constructed of blood was spotted on the card and then treated as previously
with the six data-points at each level collected in two days is cyclically described. Whenever the effective drug concentration exceeded the
used to back-calculate the concentration relative to the data-points of calibration range, the samples were diluted to fit the quantitation range
the third day. The results are then averaged. considered in the curve. The results obtained from the present method
Repeatability was independently assessed for the three days of based on DBS and our standard procedure on venous blood were then
analysis. For each day of validation, the calibration model was used to compared to evaluate the performance of the new procedure in identi­
calculate the concentrations of the three experimental replicates; then, fying and quantifying some of the analytes presented in Table 1.
the coefficient of variation (CV%) was determined for each concentra­
tion level by averaging the precision obtained for the three days. The 7. Results and discussion
intermediate precision followed a similar procedure, that make use of
the nine replications collected during all three days. In practice, the The present method proved adequate for the individual detection of
protocol used for calculating trueness and precision is based on the data all 132 target analytes and 13 internal standards at concentrations equal
collected for the 9 calibration curves, obtained in the three days [24]. or lower than 10 ng/mL. In this work, a C18 column was used to obtain
the best separation of the investigated compounds. The chromato­
5.4. Matrix effect and extraction recovery graphic run was completed in 10 min, including the final re-
equilibration time, in agreement with the efficiency requirement
The matrix effect was estimated at the concentration level of 10 ng/ needed for routine application. As shown in Table 1, 90 out of 132
mL by comparing the experimental results obtained from three blank compounds elute in the first 4.5 min, while 38 of the remaining analytes
whole blood samples and MeOH solutions, equally spiked after the (mostly synthetic cannabinoids) elute between 6.5 and 7.75 min. Even
extraction step [25]. The ionization suppression/enhancement for each when co-elution of chromatographic peaks was observed, high resolu­
target analyte was expressed as the mean percentage ratio between the tion mass spectrometry guaranteed separate quantifications of the coe­
two measured signals. Owing to coelution, several spiked analytes are luting substances by means of their differences in precursor ion and
potentially present within a single ESI droplet, resulting in the possi­ characteristic fragment ion m/z values (mass error accepted <5 ng/mL).
bility that the matrix effect may be partially due to their compresence, A typical example of extracted ion chromatogram is presented in Fig. 1,
which is unlikely in a real blood sample. For this reason, the matrix in which the group of 24 synthetic opioids - spiked in the whole blood at
effect for the compounds whose ion suppression exceeded 50% was 10 ng/mL concentration - is represented.
tested again by spiking the blank blood with the single substance. The complete results of the validation experiments for DBS samples
The extraction recovery was determined by comparing the experi­ fortified with 132 analytes at six concentrations levels (5, 7.5, 10, 25, 50
mental results obtained from three whole blood samples spiked at the and 100 ng/mL) and 13 internal standards are reported in the Supple­
concentration level of 100 ng/mL, before and after the extraction step. It mentary Tables S2-S7. Table S2 reports the outcome of the calibration

3
M. Massano et al. Talanta 241 (2022) 123265

Table 1
List of the 132 NPS/NSO substances under study (target analytes).
Compounds Formula Precursor theoretical Mass Fragment theoretical Mass Rt ISTD
5-CHLORO-AB-PINACA C18H25ClN4O2 365.17388 249.0776 5.07 JWH-081 D9
5-CHLORO-THJ018 C23H21ClN2O 377.14152 249.0771 7.12 JWH-081 D9
5-F-AB-PINACA C18H25FN4O2 349.20343 233.1066 4.61 JWH-081 D9
5-F-APINACA C23H30FN3O 384.24457 135.1168 7.18 JWH-081 D9
5-F-APP-PICA C23H26FN3O2 396.20818 232.1111 4.93 JWH-081 D9
5-F-APP-PINACA C22H25FN4O2 397.20343 233.1065 5.06 JWH-081 D9
5-F-CUMYL-PINACA C22H26FN3O 368.21327 233.1064 6.33 JWH-081 D9
5-F-NNEI MAPHTHYL isomer C24H23FN2O 375.18672 232.1112 6.42 JWH-081 D9
5-F-PB22 C23H21FN2O2 377.16598 232.1113 6.09 JWH-081 D9
AB-CHMINACA C20H28N4O2 357.2285 241.1333 5.63 JWH-081 D9
AB-FUBINACA C20H21FN4O2 369.17213 253.075 4.90 JWH-081 D9
AB-PINACA C18H26N4O2 331.21285 215.1181 5.18 JWH-081 D9
ADBICA C20H29N3O2 344.23325 214.121 5.41 JWH-081 D9
ADB-PINACA C19H28N4O2 345.2285 215.1163 5.55 JWH-081 D9
AM-1220 C26H26N2O 383.21179 98.0958 4.14 JWH-081 D9
AM-2201 C24H22FNO 360.17582 155.0488 6.49 JWH-081 D9
AM-2233 C22H23IN2O 459.09279 98.0963 3.77 JWH-081 D9
AM-694 C20H19FINO 436.05682 230.9297 6.12 JWH-081 D9
APP-FUBINACA C24H21FN4O2 417.17213 253.0755 5.23 JWH-081 D9
CB-13 C26H24O2 369.18491 155.049 7.37 JWH-081 D9
CUMYL PEGACLONE C25H28N2O 373.22744 255.1475 6.66 JWH-081 D9
JWH-007 C25H25NO 356.20089 155.0483 7.30 JWH-081 D9
JWH-015 C23H21NO 328.16959 155.0493 6.48 JWH-081 D9
JWH-018 C24H23NO 342.18524 155.0493 6.84 JWH-081 D9
JWH-019 C25H25NO 356.20089 169.0651 7.46 JWH-081 D9
JWH-020 C26H27NO 370.21654 155.0483 7.75 JWH-081 D9
JWH-073 C23H21NO 328.16959 155.0487 6.73 JWH-081 D9
JWH-081 C25H25NO2 372.19581 185.0593 7.15 JWH-081 D9
JWH-098 C26H27NO2 386.21146 155.0481 7.32 JWH-081 D9
JWH-122 C25H25NO 356.20089 169.0644 7.46 JWH-081 D9
JWH-147 C27H27NO 382.21654 155.0481 7.79 JWH-081 D9
JWH-203 C21H22ClNO 340.14627 125.015 7.01 JWH-081 D9
JWH-210 C26H27NO 370.21654 183.0802 7.57 JWH-081 D9
JWH-250 C22H25NO2 336.19581 121.0643 6.50 JWH-081 D9
JWH-251 C22H25NO 320.20089 214.1219 6.85 JWH-081 D9
JWH-302 C22H25NO2 336.19581 214.1209 6.50 JWH-081 D9
JWH-307 C26H24FNO 386.19147 155.0486 7.29 JWH-081 D9
JWH-398 C24H22ClNO 376.14627 189.0102 7.68 JWH-081 D9
MAB-CHMINACA C21H30N4O2 371.24415 241.1313 5.90 JWH-081 D9
MAM-2201 C25H24FNO 374.19147 169.0646 6.64 JWH-081 D9
MDMB-CHMINACA C22H31N3O3 386.24382 241.1324 7.22 JWH-081 D9
MDMB-CHMICA C23H32N2O3 385.24857 240.1382 6.75 JWH-081 D9
MMB-2201 C20H27FN2O3 363.20785 232.1116 5.61 JWH-081 D9
PB-22 C23H22N2O2 359.1754 214.1203 6.72 JWH-081 D9
RCS-4 C21H23NO2 322.18016 135.0439 6.50 JWH-081 D9
RCS-8 C25H29NO2 376.22711 121.0644 7.46 JWH-081 D9
STS-135 C24H31FN2O 383.24932 135.1157 7.17 JWH-081 D9
UR-144 C21H29NO 312.23219 125.0954 7.43 JWH-081 D9
UR-144-5-OH C21H29NO2 328.22711 125.0957 5.88 JWH-081 D9
WIN-48 C23H26N2O3 379.20162 135.0429 3.75 JWH-081 D9
WIN-55 C27H26N2O3 427.20162 155.0482 5.56 JWH-081 D9
XLR-11 C21H28FNO 330.22277 125.0944 6.79 JWH-081 D9
25B–NBOMe C18H22BrNO3 380.08558 121.0651 3.94 MDPV D8
25C–NBOMe C18H22ClNO3 336.1361 121.0657 4.02 25I–NBOMe D3
25H–NBOMe C18H23NO3 302.17507 121.0654 3.59 25I–NBOMe D3
25I–NBOMe C18H22INO3 428.07172 121.0639 4.10 25I–NBOMe D3
2C–B C10H14BrNO2 260.02807 243.0025 2.80 MDPV D8
2C–P C13H21NO2 224.16451 207.1355 3.45 mCPP D8
3-4 DMMC C12H17NO 192.13829 159.103 2.73 MDPV D8
4-F-METCAT C10H12FNO 182.09757 149.0631 1.66 Mephedrone D3
4-MEC C12H17NO 192.13829 174.1258 2.47 MDPV D8
5-EAPB C13H17NO 204.13829 131.0479 2.74 MDPV D8
6-APB C11H13NO 176.10699 131.0492 2.38 MDPV D8
ALFA-PVP C15H21NO 232.16959 91.0535 2.91 MDPV D8
Buphedrone C11H15NO 178.12264 131.0700 1.83 MDPV D8
Butylone C12H15NO3 222.11247 174.0917 2.25 MDPV D8
Ethylone C12H15NO3 222.11247 174.0916 2.13 PCP D5
Ketamine C13H16ClNO 238.09932 125.0149 2.42 MDPV D8
mCPP C10H13ClN2 197.084 154.0414 2.70 mCPP D8
MDPV C16H21NO3 276.15942 126.1278 2.90 MDPV D8
Mephedrone C11H15NO 178.12264 144.0810 2.27 Mephedrone D3
Methedrone C11H15NO2 194.11756 146.0602 2.09 MDPV D8
Methylone C11H13NO3 208.09682 160.0763 1.78 MDPV D8
MXE C15H21NO2 248.16451 203.1058 2.64 MDPV D8
N-Ethylcathinone C11H15NO 178.12264 132.0808 1.83 MDPV D8
(continued on next page)

4
M. Massano et al. Talanta 241 (2022) 123265

Table 1 (continued )
N-ethylpentylone C14H19NO3 250.14377 202.1211 2.79 MDPV D8
PCP C17H25N 244.20598 91.0534 3.27 PCP D5
PCP-4MEO C18H27NO 274.21654 121.0638 3.41 PCP D5
Pentedrone C12H17NO 192.13829 145.0883 2.50 Mephedrone D3
Pentylone C13H17NO3 236.12812 188.1065 2.70 MDPV D8
PMA C10H15NO 166.12264 121.0641 2.10 MDPV D8
PMMA C11H17NO 180.13829 121.0651 2.18 MDPV D8
Ritalinic acid C13H17NO2 220.13321 84.0806 2.30 MDPV D8
Trazodone C19H22ClN5O 372.15856 176.0804 3.30 MDPV D8
4-ANPP C19H24N2 281.20123 105.0689 3.44 Fentanyl D5
4-Fluorobutyrfentanyl C23H29FN2O 369.23367 188.1434 3.98 Fentanyl D5
4-methylfentanyl C23H30N2O 351.24309 202.1577 3.71 Fentanyl D5
3-methylnorfentanyl C15H22N2O 246.1732 69.6693 2.81 Fentanyl D5
Acetyl Fentanyl C21H26N2O 323.21179 188.1434 3.10 Fentanyl D5
Acrylfentanyl C22H26N2O 335.21179 188.1417 3.35 Fentanyl D5
AH-7921 C16H22Cl2N2O 329.1182 172.9545 3.50 Fentanyl D5
Alfentanil C21H32N6O3 417.26087 268.1776 3.49 Fentanyl D5
Butyrfentanyl C23H30N2O 351.24309 188.1417 3.69 Fentanyl D5
Butyryl Fentanyl (carboxy metabolite) C23H28N2O3 381.21727 188.1417 3.04 Fentanyl D5
Carfentanyl C24H30N2O3 395.23292 113.059 3.65 Fentanyl D5
Cyclopropylfentanyl C23H28N2O 349.22744 188.1419 3.54 Fentanyl D5
Despropionyl p-Fluorofentanyl C19H23FN2 299.1918 105.0691 3.60 Fentanyl D5
Fentanyl C22H28N2O 337.22744 188.1434 3.55 Fentanyl D5
FuranilFentanil C24H26N2O2 375.2067 188.1434 3.67 Fentanyl D5
Hydrocodone C18H21NO3 300.15942 199.0761 2.16 Fentanyl D5
MT-45 C24H32N2 349.26383 181.0999 4.12 Fentanyl D5
OCFENTANYL C22H27FN2O2 371.21293 188.1418 3.25 Fentanyl D5
Remifentanil C20H28N2O5 377.2071 113.0603 2.92 Fentanyl D5
Sufentanil C22H30N2O2S 387.21008 238.1268 3.82 Fentanyl D5
Tramadol C16H25NO2 264.19581 58.0648 2.71 Fentanyl D5
U-47700 C16H24Cl2N2O 331.13385 284.061 3.42 Fentanyl D5
Valerylfentanyl (carboxy metabolite) C24H30N2O3 395.23292 113.0589 3.65 Fentanyl D5
4-ACETOXY DIPT C18H26N2O2 303.2067 160.0747 3.10 MDPV D8
4-ACETOXY DMT C14H18N2O2 247.1441 174.0899 2.58 MDPV D8
5-MeO-DALT C17H22N2O 271.18049 174.0912 3.10 MDPV D8
5-MeO-DIPT C17H26N2O 275.21179 174.091 3.31 MDPV D8
5-METOXY AMT C12H16N2O 205.13354 147.0672 2.25 PCP D5
DMT C12H16N2 189.13862 2.06 Mescaline D9
HARMINE C13H12N2O 213.10224 170.0826 2.88 MDPV D8
LSD C20H25N3O 324.20704 223.1227 3.13 MDPV D8
Mescaline C11H17NO3 212.12812 195.1019 1.98 PCP D5
Mitragynine C23H30N2O4 399.22783 174.0914 3.69 MDPV D8
Psilocin C12H16N2O 205.13354 132.0808 1.78 MDPV D8
Norfentanyl C14H20N2O 233.16484 144.0808 2.46 Norfentayl D5
5-F-ADB C20H28FN3O3 378.21875 233.1061 6.17 JWH-081 D9
ADB-FUBINACA C21H23FN4O2 383.18778 253.0747 5.19 JWH-081 D9
β-Phenylfentanyl C28H32N2O 413.25874 188.1398 4.27 Fentanyl D5
Butyrilnorfentanyl C15H22N2O 247.18049 84.0800 3.00 Fentanyl D5
Ethylphenidate C15H21NO2 248.16451 84.0797 3.05 MDPV D8
Ethyltryptamine C12H16N2 189.13862 130.0643 2.65 MDPV D8
Furanylnorfentanyl C16H18N2O2 271.1441 84.0798 2.55 Fentanyl D5
Hydroxy-fentanyl C22H28N2O2 353.22235 204.135 3.11 Fentanyl D5
Hydroxy-thiofentanyl C20H26N2O2S 359.17878 192.0829 2.94 Fentanyl D5
Fentanyl C26H28N2O 385.22744 188.1434 3.91 Fentanyl D5
5-MeO-DMT C13H18N2O 219.14919 58.0691 2.22 MDPV D8
Phenylacetylfentanyl C27H30N2O 399.24309 188.1408 3.05 Fentanyl D5
Eutylone C13H17NO3 236.12812 188.1060 2.64 MDPV D8
PCP-D5 C17H20[2H]5N 249.23736 96.0854 3.24
AH-7921 D6 C16H1[2H]6Cl2N2O 335.15586 172.9546 3.64
Norfentayl D5 C14H15[2H]5N2O 238.19622 84.0800 2.40
Oxycodone D6 C18H15[2H]6NO4 322.192 304.1792 1.93
Fentanyl D5 C22H23[2H]5N2O 342.25882 188.1416 3.50
JWH-250 D5 C22H20[2H]5NO2 341.22719 121.0637 6.73
JWH-081 D9 C25H16[2H]9NO2 381.2523 185.0577 7.24
JWH-018 D9 C24H14[2H]9NO 351.24173 155.0485 7.03
Mephedrone D3 C11H12[2H]3NO 181.14147 148.1064 2.35
mCPP D8 C10H5[2H]8ClN2 205.13422 158.0661 2.65
25I-nBOMe D3 C18H19[2H]3INO3 431.09055 124.0835 4.09
Mescaline D9 C11H8[2H]9NO3 221.18461 171.0921 1.90
MDPV D8 C16H13[2H]8NO3 284.20963 134.1778 2.97

process. Analysis of residues and variances of calibration data points at statistically significant for all but one (JWH-015) tested substances.
low, medium and high concentration levels showed that heteroscedastic Consequently, a quadratic calibration model was chosen for all analytes.
distributions were present for all the target analytes, making the intro­ The LOD values, calculated using the corrected Hubaux-Vos algo­
duction of weighting factors in the calibration (either 1/x of 1/x2) rithm, ranged from 1.3 ng/mL for ethylphenidate and 1.8 ng/mL for 4-F-
beneficial. Moreover, the quadratic term of the calibration model proved butyrylfentanyl up to 6.3 ng/mL for UR-144. The LOD was then verified

5
M. Massano et al. Talanta 241 (2022) 123265

Fig. 1. Chromatographic profile of 24 synthetic opioids included in the panel within a 1.9–4.3 min retention time interval. Extracted ion chromatograms (XICs)
resulting from the optimized data acquisition method, obtained from the 10 ng/mL neat standard mixture. Method was built using the Scheduled Algorithm Pro in
SCIEX OS Software. In the figure: 1) Hydrocodone, 2) Norfentanyl, 3) Butyryl norfentanyl, 4) Remifentanil, 5) Butyryl fentanyl carboxy metabolite, 6) Acetyl
Fentanyl, 7) Ocfentanyl, 8) Carfentanyl, 9) Alfentanil, 10) Acrylfentanyl, 11) U-47700, 12) 4-ANPP, 13) Fentanyl, 14) Despropionyl p-fluorofentanyl, 15) AH-7921,
16) FuranilFentanil, 17) Butyrylfentanyl, 18) Cyclopropylfentanyl, 19) Valeryl fentanyl carboxy metabolite 20) 4-F-Butyrylfentanyl, 21) 4-Methyl fentanyl, 22)
Tramadol, 23) Sufentanil, 24) MT-45.

by spiking blank blood samples with decreasing concentrations (7.5, 5, showed CV% in the range 15%–28%).
2 ng/mL) until a response equivalent to three times the background Further tests were performed in order to evaluate a possible corre­
noise was observed. This verification process proved that for 74 analytes lation between the recovery yield of fentanyl, norfentanyl and ketamine
out of 132 (56%), a 2 ng/mL LOD was experimentally verified, namely a and their compresence, as they are frequently present in the same real
lower LOD value than the calculated ones (Table S3). This evidence sample. The results showed (Table S5) that the extraction recovery did
suggests that the Hubaux-Vos algorithm corrected with Currie’s method not vary significantly, no matter if only these three substances were
provide reliable yet conservative LOD values. spiked in the blank blood or all 132 analytes were simultaneously pre­
The experimental results obtained for LOD values confirmed that the sent in the sample.
overall method sensitivity is adequate for the detection of NPS extracted Intermediate precision and repeatability (expressed as percent vari­
from DBS in routine applications. In fact, the 2 ng/mL limit consistently ation coefficient, CV%) and trueness (expressed as bias %) for all 132
assessed in the present study represents the actual lower LOD measured target analytes are reported in Table S3. Notably, the validation pro­
in several other studies [28–30] and an acceptable requirement for NPS cedure adopted in the present study allows precise and accurate calcu­
detection in blood, based on information available from case reports lation at all concentrations involved in the calibration process (6
[31–33]. On the other hand, specific recommendations for DBS detec­ calibration levels), not only at low, intermediate and high concentra­
tion in blood have not been established yet, as long as little clinical in­ tions, as most recommended validation protocols entail. An example is
formation is available at the moment. reported in Fig. 2, showing precision and trueness trends for AH-7921.
The extraction recovery was found to depend on the different classes The repeatability and intermediate precision and trueness trend for all
of drugs considered (Figure S4), being higher for the class of synthetic the classes of compounds studied is reported in the Supplementary
cannabinoids (range 50%–100%) particularly for the JWH-series, and Figure S6.
lower for synthetic cathinones and hallucinogens (range 30%–60%) and In particular, the aggregate inter-day trueness data are reported in
fentanyl analogues and synthetic opioids (range 30–50%). Relatively Fig. 3, showing extremely limited deviation from the expected concen­
low extraction recoveries are expected from DBS cards [34], also taking trations (− 10% < bias% < +10%) for 114 target analytes out of 132.
into account that a large number of analytes have to be extracted The inter-day trueness is an especially important performance param­
simultaneously from the same dried droplet. However, the extraction eter because it provides a reasonable estimation of the quantitative
repeatability measured from three replications proved to be satisfactory measurement reliability under routine conditions and – indirectly – the
for the large majority of substances (CV% lower than 15% for 113 out of quality of calibration. This positive outcome partly relies on its aver­
132) [26], with the peculiar exception of the JWH-series that combines aging the single results of repeated determinations, forasmuch as the
high extraction yield with relatively lower repeatability (10 out of 17 random sources of variability that influence the precision results

Fig. 2. Intermediate precision and repeatability (CV%) and trueness (bias%) trends for the synthetic opioid AH-7921 at the different calibration levels.

6
M. Massano et al. Talanta 241 (2022) 123265

Fig. 3. Inter-day trueness (bias%) trends for the class of a) synthetic cannabinoids, b) cathinones and hallucinogens and c) synthetic opioids.

(extraction yield, matrix effect, etc.) may find equalization. Therefore, it limits observed under stressed conditions rather than expected values in
is highly recommended to collect blood droplets on three separate DBS real toxicological contexts. Again, whenever accurate quantitative
cards and average their results whenever accurate quantitative de­ determination of a single substance is needed, specific experiments
terminations are needed. should be planned to complete the validation for that specific substance
Precision data, described by CV% from repeated independent ana­ and an on-purpose calibration model may be prudentially built, so as to
lyses, show rather homogeneous results within each class and between avoid overestimation of the matrix effect.
different classes of substances under study (Supplementary Table S3 and The largest part of the target analytes showed good stability on DBS
Figure S6). Therefore, the observed data variability can realistically be cards after drying, even for long periods of time. At both concentration
attributed to the analytical method itself rather than to specific chemical levels tested, deviations from the nominal value fell within 15% for most
properties of the investigated substances or their interaction with the of them. Quite obviously, the NPS stability proved to be slightly more
DBS card. Quite interestingly, the average intermediate precision result affected when a storage temperature of 35 ◦ C was maintained, in com­
(CV% = 15.5%, N = 132) is only slightly higher than the intra-day bination with a high storage period and low analyte concentration.
variability (CV% = 11.4%, N = 132), as is expected for a substantially Fig. 4 reports an example of this trend. In particular, synthetic canna­
stable method and calibration. binoids – and especially the JWH group – showed lower stability than
Electrospray ionization produced moderate ion suppression, result­ the other classes of NPS. An example of particularly poor stability is
ing in matrix effect values ranging between negligible to − 50% for evident in Fig. 5 for JWH-015: on one hand, acceptable conservation
synthetic cannabinoids (average − 29%), between negligible to − 40% (>90%) is guaranteed during the initial 7 days at any storage temper­
for synthetic cathinones and hallucinogens (average − 24%) and fenta­ ature, on the other hand it is evident that significant degradation
nyl analogues and synthetic opioids (average − 27%) (Table S7). Due to occurred after 40 storage days even if the DBS cards were maintained at
the presence of many spiked analytes inside the single blood drop and − 20 ◦ C. In summary, the stability data suggest that storage of the DBS
the crowding of peaks around certain retention times (for example, in cards at − 20 ◦ C is not essential, the most convenient storage condition
the interval 3.4–4.0 min), we considered the possibility that part of the apparently being at 4 ◦ C. As a matter of fact, extremely limited degra­
matrix effect could be attributed to their coelution and interaction. To dation is observed at 4 ◦ C for periods fully compatible with routine
test this hypothesis, we measured again the matrix effect using blood analytical processing, while an acceptable level of preservation is
droplets spiked with only the five substances whose ion suppression maintained up to 40 days for most NPS. Specific caution should be
value exceeded − 50% in the first set of experiments. Indeed, ion sup­ exercised in counter-analyses involving the confirmation of JWH-series
pression decreased for all target analytes, on average from − 62% to positive testing. Further studies will be needed to evaluate the com­
− 29%. In particular, the effect reduction for 5-chloro-TH-J018 was from pounds stability in real samples, similarly to what has been done on
− 69% to − 9%, for AM-2201 from − 57% to − 43%, for JWH-007 from post-mortem samples for psychoactive substances [11]. Indeed, spiked
− 66% to − 7%, for MMB-2201 from − 58% to − 38%, and for WIN-48 samples may not always display the same stability profile as real
from − 59% to − 49%. These results suggest that the recorded ion sup­ samples.
pression initially attributed to the blood matrix is likely to be influenced The proven stability of NPS on DBS cards, combined with ease of
by the coelution of several spiked analytes. In real samples, the simul­ sampling and minimal storage volume, represents a key asset of the DBS
taneous presence of more than five drugs in a single blood sample is technique confirming its use as an alternative and innovative sampling
quite implausible to occur, even in the worst cases. In conclusion, the ion method in troublesome conditions, for example when the sampling site
suppression data reported in Table S6 should be considered as upper is far from the laboratory and/or few days are required after sampling

7
M. Massano et al. Talanta 241 (2022) 123265

Fig. 4. Stability of 25C-NBoMe at different temperature conditions (− 20, 4, 35 ◦ C) during 1, 3, 7, 40 days.

Fig. 5. Stability of JWH-015 at different temperature conditions (− 20, 4, 35 ◦ C) during 1, 3, 7, 40 days.

until the analysis is made possible. third parties, but the progressive introduction of NPS in the illegal
market significantly complicated these tests and expanded the number
8. Application to real samples of targeted analytes. It is nonetheless essential to discriminate the sub­
ject who are under the effect of drugs from those who may have
After completing the validation, the present method was applied to consumed drugs days earlier, i.e. out of the control context. The matrix
seven venous whole blood samples which previously tested positive to of choice for this discrimination is blood, but blood sampling is pre­
ketamine and fentanyl in our laboratory. The results summarized in vented in almost all circumstances by ethical and practical reasons.
Table 2 show that the new procedure based on DBS sample collection The present method overcomes both problems by combining the use
allowed confirmation of the ketamine and fentanyl positive testing, of UHPLC-QTOF-HRMS instrumentation with a simple and minimally
whenever their concentrations was higher than the corresponding LOQ invasive DBS sample collection for detecting as many as 132 traditional
of the method. Also, the quantitative results proved consistent with drugs and NPS selected from EMCDDA reports [35] and articles [36,37],
those measured with the routine procedure based on large blood volume together with their metabolites, on only 30 μL whole blood. The
sampling. These results demonstrate the ability of the new technique to analytical method’s validation confirmed its reliability for the extraction
detect and correctly quantify the substances present in the sample. and accurate analysis of this wide array of structurally different NPS
within an adequate concentration interval (typically, 5–100 ng/mL).
9. Conclusions The initial comparison on real toxicology samples between our tradi­
tional routine procedure based on high volume blood sampling and the
The need to test biological samples for drugs of abuse keeps new DBS procedure provides a preliminary confirmation of the potential
increasing in several social contexts, especially in workplace and road applicability of the latter technique on a vast scale in several fields,
controls, so as to guarantee safer conditions for workers, drivers, and including workplace drug testing, road controls, and drug monitoring in

Table 2
Ketamine and fentanyl concentrations determined on venous whole blood (ng/mL) from real toxicology casework and measured with both the routine laboratory
procedure and the present DBS method. Values between parentheses are rough estimations below LOQ; n.d. = not detected.
ID Sample Routine method (UHPLC–ESI-MS/MS) DBS method (UHPLC-QTOF-HRMS)
Ketamine (ng/mL) Fentanyl (ng/mL) Ketamine (ng/mL) Fentanyl (ng/mL)
S1 n.d. 0.3 n.d. < LOQ
S2 37.5 n.d. 42.0 n.d.
S3 493 n.d. 597 n.d.
S4 447 0.2 600 n.d.
S5 74.0 2.0 88.0 < LOQ (~2.3)
S6 405 n.d. 700 n.d.
S7 275 2.1 140 < LOQ (~2.5)

8
M. Massano et al. Talanta 241 (2022) 123265

clinical and forensic contexts. [12] M. Thevis, T. Kuuranne, A. Thomas, H. Geyer, Do dried blood spots have the
potential to support result management processes in routine sports drug testing? -
Prospectively, the new method will be applied on real capillary
Part 2: proactive sampling for follow-up investigations concerning atypical or
whole blood samples in comparison with venous blood. Further inves­ adverse analytical findings, Drug Test. Anal. 13 (2021) 505–509, https://doi.org/
tigation will be devoted to the hematocrit effect (HT) and its influence 10.1002/dta.3011.
on the volumetric blood sampling method, either a single drop deposited [13] A. Deep, P. Kumar, A. Kumar, A. Thakkar, Dry blood spot technique: a review, Int.
J. Pharmaceut. Sci. Rev. Res. 15 (2012) 90–94.
on the card (requiring a corrective factor for the HT effect) or other [14] N. Sadones, S. Capiau, P.M.M. De Kesel, W.E. Lambert, C.P. Stove, Spot them in the
micro-sampling systems (VAMS). spot: analysis of abused substances using dried blood spots, Bioanalysis 6 (2014)
2211–2227, https://doi.org/10.4155/bio.14.156.
[15] H.Y. Tey, H.H. See, A review of recent advances in microsampling techniques of
Funding biological fluids for therapeutic drug monitoring, J. Chromatogr., A 1635 (2021),
461731, https://doi.org/10.1016/j.chroma.2020.461731.
This research did not receive any specific grant from funding [16] R.V. Oliveira, J. Henion, E. Wickremsinhe, Fully-automated approach for online
dried blood spot extraction and bioanalysis by two-dimensional-liquid
agencies in the public, commercial, or not-for-profit sectors. chromatography coupled with high-resolution quadrupole time-of-flight mass
spectrometry, Anal. Chem. 86 (2014) 1246–1253, https://doi.org/10.1021/
Credit author statement ac403672u.
[17] W. Harry Hannon, B.L. Therrell, Overview of the history and applications of dried
blood samples, in: Dried Blood Spots, Wiley-Blackwell, 2014, pp. 1–15, https://doi.
Marta Massano: Conceptualization, Data curation, Methodology, org/10.1002/9781118890837.ch1.
Project administration, Writing – original draft. Carola Incardona: [18] E. Alladio, E. Amante, C. Bozzolino, F. Seganti, A. Salomone, M. Vincenti,
B. Desharnais, Effective validation of chromatographic analytical methods: the
Conceptualization, Data curation. Enrico Gerace: Conceptualization,
illustrative case of androgenic steroids, Talanta 215 (2020), 120867, https://doi.
Data curation, Methodology, Writing – review & editing. Pierre Negri: org/10.1016/j.talanta.2020.120867.
Software, Supervision. Eugenio Alladio:Validation, Methodology. [19] B. Desharnais, F. Camirand-Lemyre, P. Mireault, C.D. Skinner, Procedure for the
selection and validation of a calibration model i-description and application,
Alberto Salomone: Conceptualization, Methodology, Project adminis­
J. Anal. Toxicol. 41 (2017) 261–268, https://doi.org/10.1093/jat/bkx001.
tration, Writing – review & editing. Marco Vincenti: Data curation, [20] E. Alladio, E. Amante, C. Bozzolino, F. Seganti, A. Salomone, M. Vincenti,
Writing – review & editing B. Desharnais, Experimental and statistical protocol for the effective validation of
chromatographic analytical methods, MethodsX 7 (2020), 100919, https://doi.
org/10.1016/j.mex.2020.100919.
Declaration of competing interest [21] H. Gu, G. Liu, J. Wang, A.-F. Aubry, M.E. Arnold, Selecting the correct weighting
factors for linear and quadratic calibration curves with least-squares regression
The authors declare that they have no known competing financial algorithm in bioanalytical LC-MS/MS assays and impacts of using incorrect
weighting factors on curve stability, data quality, and assay performance, Anal.
interests or personal relationships that could have appeared to influence Chem. 86 (2014) 8959–8966, https://doi.org/10.1021/ac5018265.
the work reported in this paper. [22] Andre Hubaux, Gilbert Vos, Decision and detection limits for calibration curves,
Anal. Chem. 42 (1970) 849–855, https://doi.org/10.1021/ac60290a013.
[23] L.A. Currie, Detection and quantification limits: origins and historical overview,
Appendix A. Supplementary data Anal. Chim. Acta 391 (1999) 127–134, https://doi.org/10.1016/S0003-2670(99)
00105-1.
Supplementary data to this article can be found online at https://doi. [24] B. Desharnais, F. Camirand-Lemyre, P. Mireault, C.D. Skinner, Procedure for the
selection and validation of a calibration model II-Theoretical basis, J. Anal.
org/10.1016/j.talanta.2022.123265. Toxicol. 41 (2017) 269–276, https://doi.org/10.1093/jat/bkx002.
[25] Scientific working group for forensic toxicology, scientific working group for
References forensic toxicology (SWGTOX) standard practices for method validation in forensic
toxicology, J. Anal. Toxicol. 37 (2013) 452–474, https://doi.org/10.1093/jat/
bkt054.
[1] A. Peacock, R. Bruno, N. Gisev, L. Degenhardt, W. Hall, R. Sedefov, J. White, K.
[26] S. Capiau, H. Veenhof, R.A. Koster, Y. Bergqvist, M. Boettcher, O. Halmingh, B.
V. Thomas, M. Farrell, P. Griffiths, New psychoactive substances: challenges for
G. Keevil, B.C.P. Koch, R. Linden, C. Pistos, L.M. Stolk, D.J. Touw, C.P. Stove, J.-W.
drug surveillance, control, and public health responses, Lancet 394 (2019)
C. Alffenaar, Official international association for therapeutic drug monitoring and
1668–1684, https://doi.org/10.1016/S0140-6736(19)32231-7.
clinical toxicology guideline: development and validation of dried blood
[2] D. Baumeister, L.M. Tojo, D.K. Tracy, Legal highs: staying on top of the flood of
spot–based methods for therapeutic drug monitoring, Ther. Drug Monit. 41 (2019)
novel psychoactive substances, Ther. Adv. Psychopharmacol. 5 (2015) 97–132,
409–430, https://doi.org/10.1097/FTD.0000000000000643.
https://doi.org/10.1177/2045125314559539.
[27] M. Vincenti, D. Cavanna, E. Gerace, V. Pirro, M. Petrarulo, D. Di Corcia,
[3] S.M. Aarde, M.A. Taffe, Predicting the abuse liability of entactogen-class, new and
A. Salomone, Fast screening of 88 pharmaceutical drugs and metabolites in whole
emerging psychoactive substances via preclinical models of drug self-
blood by ultrahigh-performance liquid chromatography–tandem mass
administration, Curr. Top. Behav. Neurosci. 32 (2017) 145–164, https://doi.org/
spectrometry, Anal. Bioanal. Chem. 405 (2013) 863–879, https://doi.org/
10.1007/7854_2016_54.
10.1007/s00216-012-6403-y.
[4] A. Shafi, A.J. Berry, H. Sumnall, D.M. Wood, D.K. Tracy, New psychoactive
[28] Y. Wang, Y. Shi, Y. Yu, L. Chen, J. Jiang, J. Long, P. Xiang, G. Duan, Screening of
substances: a review and updates, Ther. Adv. Psychopharmacol. 10 (2020), https://
synthetic cathinones and metabolites in dried blood spots by UPLC–MS-MS,
doi.org/10.1177/2045125320967197, 2045125320967197.
J. Anal. Toxicol. 45 (2020), https://doi.org/10.1093/jat/bkaa106.
[5] Observatoire européen des drogues et des toxicomanies, European Drug Report:
[29] S. Sadler Simões, A. Castañera Ajenjo, M.J. Dias, Dried blood spots combined to an
Trends and Developments, Publications office of the European Union, Luxembourg,
UPLC-MS/MS method for the simultaneous determination of drugs of abuse in
2021.
forensic toxicology, J. Pharm. Biomed. Anal. 147 (2018) 634–644, https://doi.org/
[6] A. Salomone, G. Gazzilli, D. Di Corcia, E. Gerace, M. Vincenti, Determination of
10.1016/j.jpba.2017.02.046.
cathinones and other stimulant, psychedelic, and dissociative designer drugs in real
[30] M. Concheiro, S. Anizan, K. Ellefsen, M.A. Huestis, Simultaneous quantification of
hair samples, Anal. Bioanal. Chem. 408 (2016) 2035–2042, https://doi.org/
28 synthetic cathinones and metabolites in urine by liquid chromatography-high
10.1007/s00216-015-9247-4.
resolution mass spectrometry, Anal. Bioanal. Chem. 405 (2013) 9437–9448,
[7] H.R. Sumnall, M. Evans-Brown, J. McVeigh, Social, policy, and public health
https://doi.org/10.1007/s00216-013-7386-z.
perspectives on new psychoactive substances, Drug Test. Anal. 3 (2011) 515–523,
[31] J. Welter-Luedeke, H.H. Maurer, Relevance of published blood concentrations of
https://doi.org/10.1002/dta.310.
new psychoactive substance for rational case interpretation, WIREs Forensic Sci. 1
[8] C.C. Scriver, A simple phenylalanine method for detecting phenylketonuria in large
(2019) e1174, https://doi.org/10.1002/wfs2.1174.
populations of newborn infants, by Robert Guthrie and Ada Susi, Pediatrics 32
[32] K. Shintani-Ishida, K. Saka, M. Nakamura, K.-I. Yoshida, H. Ikegaya, Experimental
(1963) 318–343. Pediatrics 102 (1998) 236–237.
study on the postmortem redistribution of the substituted phenethylamine, 25B-
[9] K. Malsagova, A. Kopylov, A. Stepanov, T. Butkova, A. Izotov, A. Kaysheva, Dried
NBOMe, J. Forensic Sci. 63 (2018) 588–591, https://doi.org/10.1111/1556-
blood spot in laboratory: directions and prospects, Diagn. Basel Switz. 10 (2020)
4029.13583.
E248, https://doi.org/10.3390/diagnostics10040248.
[33] M. Diestelmann, A. Zangl, I. Herrle, E. Koch, M. Graw, L.D. Paul, MDPV in forensic
[10] C.P. Stove, A.-S.M.E. Ingels, P.M.M. De Kesel, W.E. Lambert, Dried blood spots in
routine cases: psychotic and aggressive behavior in relation to plasma
toxicology: from the cradle to the grave? Crit. Rev. Toxicol. 42 (2012) 230–243,
concentrations, Forensic Sci. Int. 283 (2018) 72–84, https://doi.org/10.1016/j.
https://doi.org/10.3109/10408444.2011.650790.
forsciint.2017.12.003.
[11] M. Moretti, A. Manfredi, F. Freni, C. Previderé, A.M.M. Osculati, P. Grignani,
[34] R.A. Koster, D.J. Touw, J.-W.C. Alffenaar, Dried blood spot analysis; facing new
L. Tronconi, C. Carelli, C. Vignali, L. Morini, A comparison between two different
challenges, J. Appl. Bioanal. 1 (2015) 38–41, https://doi.org/10.17145/
dried blood substrates in determination of psychoactive substances in postmortem
jab.15.007.
samples, Forensic Toxicol. 39 (2021) 385–393, https://doi.org/10.1007/s11419-
020-00567-2.

9
M. Massano et al. Talanta 241 (2022) 123265

[35] The EU Early Warning System on new psychoactive substances (NPS) | www. Forensic Sci. Int. 326 (2021), 110904, https://doi.org/10.1016/j.
emcdda.europa.eu. https://www.emcdda.europa.eu/publications/topic-overvie forsciint.2021.110904.
ws/eu-early-warning-system. [37] J. Lyden, I.A. Binswanger, The United States opioid epidemic, Semin. Perinatol. 43
[36] F. Vincenti, A. Gregori, M. Flammini, F. Di Rosa, A. Salomone, Seizures of new (2019) 123–131, https://doi.org/10.1053/j.semperi.2019.01.001.
psychoactive substances on the Italian territory during the COVID-19 pandemic,

10