Journal of Pharmaceutical and Biomedical Analysis 106 (2015) 144–152
Contents lists available at ScienceDirect
Journal of Pharmaceutical and Biomedical Analysis journal homepage: www.elsevier.com/locate/jpba
Liquid chromatography–high resolution mass spectrometry (LC–HRMS) determination of stimulants, anorectic drugs and phosphodiesterase 5 inhibitors (PDE5I) in food supplements Sabina Strano-Rossi a,∗ , Sara Odoardi a , Erika Castrignanò a , Giovanni Serpelloni b , Marcello Chiarotti a a b
Institute of Public Health, Forensic Toxicology Laboratory, Università Cattolica del S. Cuore, Rome, Italy Antidrug Policies Department, Presidency of Council of Ministers, Rome, Italy
a r t i c l e
i n f o
Article history: Received 5 May 2014 Received in revised form 6 June 2014 Accepted 7 June 2014 Available online 14 June 2014 Keywords: LC–HRMS Nutritional supplements Stimulants PdE5Is Anorectics
a b s t r a c t The paper describes a liquid chromatography/high resolution mass spectrometry LC/HRMS method for the simultaneous identification and quantification of stimulants (ephedrines, caffeine, anorectic drugs such as phentermine, phendimetrazine, phenmetrazine, fenfluramine, benfluorex, mephentermine, fencanfamine, sibutramine) and PDE5I (sildenafil, vardenafil and tadalafil) in food supplements using a benchtop Orbitrap mass spectrometer. The mass detector, with a nominal resolving power of 100,000 (FWHM at m/z 200), operated in full scan mode in ESI positive ionization mode. Analytes were identified by retention times, accurate masses and correspondence of experimental and calculated isotopic patterns. The limits of detection (LOD) obtained varied from 1 to 25 ng g−1 and limits of quantification (LOQ) were 50 ng g−1 for all compounds. The method was linear for all the analytes in the ranges from 50 to 2000 ng g−1 , giving correlation coefficients > 0.99. Accuracy (intended as %E) and repeatability (% CV) were always lower than 15%. The method was applied to the analysis of 36 dietary supplements, revealing the presence of ephedrine and/or pseudoephedrine in four of them, caffeine in eight of them and sildenafil in four of them. In one case, ephedrine was not reported on the label of the dietary supplement, as well as for caffeine in other two cases. A further confirmation of the analytes identity in positive samples was obtained through in-source fragmentation and comparison of the obtained fragments and their relative abundances with those from certified standards. As the acquisition mode is full scan, it would be also possible to re-process a previously acquired datafile for the investigation of untargeted analytes. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Dietary supplements are concentrated sources of nutrients that are theoretically present in a normal and balanced diet, such as minerals, vitamins, amino acids, trace elements and other components whose purpose is to supplement the normal diet. They are sold in form of tablets, capsules or in powder for an oral consumption. All the layers of population are prone to intake nutritional supplements and herbal remedies, as these preparations are generally considered safer than synthetic drugs and claim to benefit consumers for treating various health conditions and promoting
∗ Corresponding author at: Institute of Public Health, Forensic Toxicology Laboratory, Università Cattolica del S. Cuore, l.go F. Vito, 1, Rome, Italy. Tel.: +39 0630154249; fax: +39 063051168. E-mail addresses: [email protected], [email protected] (S. Strano-Rossi). http://dx.doi.org/10.1016/j.jpba.2014.06.011 0731-7085/© 2014 Elsevier B.V. All rights reserved.
general wellbeing. The widespread diffusion of Internet sites and gyms selling supplements render difficult the application of international and national directives regarding to the labeling and the controls on food supplements and on pharmaceutical products [1,2]. This poses a risk to the public health and it can be considered a legal offense in many Countries. Furthermore, supplements safety is of major concern, since chemicals can be added illegally to dietary supplements in order to enhance the activities claimed (anorectic, stimulating, erectile function stimulating, diuretic). During recent years it has been reported that products marketed as dietary supplements could contain non-labeled substances, like clenbuterol [3], the anorectic drugs sibutramine [4–7], phentermine and fenfluramine [8], methylenedioxymetamphetamine (MDMA) [9], phosphodiesterase 5 inhibitors (PDE5I) like sildenafil, vardenafil, tadalafil [10–13] and analogs [13,14]. Moreover, all of the above-mentioned substances, with the exception of PDE5Is, are included in the World Anti-Doping Agency (WADA) list of prohibited substances, and their intake, also if
S. Strano-Rossi et al. / Journal of Pharmaceutical and Biomedical Analysis 106 (2015) 144–152
involuntary or accidental, can result in a doping offense for the athlete, giving rise to a positive doping test [15]. In fact not only amateur sports people but also professional athletes are tempted to take dietary supplements as they believe assuming natural and safer products in order to modify their body’s appearance, to promote fat burning and to improve their performances by a stimulating effect. Besides the risk of violating doping rules, both athletes and, more in general, all population are exposed to the high health risk involved in the intake of products that may contain unlabeled drugs with potentially dangerous unwanted effects. The analysis of food supplements, to exclude the presence of drugs, is therefore of fundamental importance for the evaluation of products’ safety, especially when they are sold by gyms as well as through not-conventional channels such as internet websites and “spice shops”. The aim of this paper is to provide a valid method for the simultaneous screening of stimulants and PDE5I in nutritional supplements, performed by Liquid Chromatography (LC) High Resolution Mass Spectrometry (HRMS) using an Exactive benchtop Orbitrap mass spectrometer, after a simple extraction of active ingredients and possible contaminants from food supplements. The main advantage of the use of Orbitrap mass spectrometry for the analysis of complex matrices is the possibility of obtaining a mass resolution till 100,000 (full width at half maximum FWHM) at m/z 200, with a mass accuracy of 2 mDa (as stated by the manufacturer), achieved by using for the internal calibration a lockmass such as the diisooctyl phtalate ionic species, normally present in a laboratory environment. This leads to a high specificity of the methodology for the target compounds, useful for the identification of contaminants in complex matrices like food supplements. This technology, in fact, achieving a high mass resolution and hence a high accuracy of mass measurement, enhances the possibility to distinguish the target analyte from nearly isobaric compounds that, if co-eluting, can interfere in the analysis, even after a prior sample purification step. Possible interferences can therefore be resolved by the use of HRMS. Another main advantage of this technique is that the analyses are performed in full scan mode. This enables a further investigation of a wider range of compounds, even if they were not included at the early stage of the method development, by means of a simple re-elaboration of the previously acquired datafile. 2. Materials and methods 2.1. Chemicals and reagents Formic acid, sodium hydroxide, ultrapure water and methanol were purchased from Sigma–Aldrich (Milano, Italy). All solvents used were LC–MS grade. Clenbuterol was from Sigma–Aldrich (Saint Louis, MO, USA). Caffeine, fenfluramine, phendimetrazine, phenmetrazine, phentermine, mephentermine, ephedrine, pseudoephedrine, norpseudoephedrine, norephedrine, sibutramine, benfluorex and methamphetamine-D5 were from LGC Standards (Milano, Italy); fencamfamine was from Salars (Como, Italy). Methanolic stock solutions (1 mg mL−1 ) of all substances were used for the preparation of the working solution mixture, containing the analytes at a concentration of 10 g mL−1 in methanol. It was stored in freezer at −20 ◦ C. 2.2. Sample preparation Nutritional supplements in powder, oily capsules and tablets seized in fitness leisure centers or sold in the internet market were analyzed in order to check the eventual presence of unlabeled substances. As declared on the labels, they contained proteins, vitamins
145
and fruit, chocolate or vanilla flavors, caffeine in six supplements (labeled as guaranà in four of them) and ephedrines (labeled as ephedrine or ma huang, in four of them), or, in the case of a product called “Kamagra”, sildenafil. Aliquots of these nutritional supplements were collected in glass tubes and kept refrigerated at 4 ◦ C in the dark in order to avoid photo-degradation. 100 mg of the sample were weighted, the internal standard (IS) methamphetamine-D5 was added to the samples at 50 ng g−1 and five milliliters of methanol were added. The specimen was then vortexed for 20 s and centrifuged for 10 min at 4000 rpm. The methanolic phase was evaporated to dryness under nitrogen flow at 40 ◦ C and reconstituted in 2 mL of a solution of sodium hydroxide 1 M. A liquid/liquid extraction was performed through the addiction of five milliliters of a mixture of pentane/ethyl ether 9:1, under agitation for an hour [16]. After centrifugation, the organic phase was evaporated to dryness under a nitrogen stream at room temperature and reconstituted with methanol/water/formic acid 6:4:0.03 (100 L) in a glass vial. 10 L were then directly injected in LC–HRMS system. 2.3. LC–HRMS 2.3.1. Equipment The LC–HRMS system was composed of a Thermo ULTIMATE 3000 system equipped with an analytical column Thermo Acclaim RSLC 120 C18 (2.1 mm × 100 mm, 2.2 m particle size), coupled to a Thermo single-stage Orbitrap (Exactive) MS system. The whole equipment and the column were provided by Thermo Fisher Scientific (Milan, Italy).
2.3.2. LC–HRMS conditions The analytical column was maintained at 40 ◦ C, while the sample manager at 15 ◦ C. The following mobile phases were used: (A) ultrapure water/methanol/formic acid 9/1/0.1, (B) methanol acidified with 0.1% formic acid. Separation was achieved using the following mobile phase gradient: 100% A for 1 min, from 0% to 10% of phase B in 0.1 min, linear gradient to 15% B in 4 min, linear gradient to 50% B in 1.8 min, to 70% B in 1.7, to 80% B in 1.1 min, to 100% in 1 min held for 3.5 min. The injection volume was 10 L and the flow rate was set at 0.4 mL min−1 . The chromatographic run time was of 14.5 min per analysis. Ionization parameters were optimized with direct infusion of each substance, including IS methamphetamine-D5, by an external syringe at 10 L min−1 . The chosen tune for the HESI source had the following settings: source current at 5 A, sheath gas and auxiliary gas (either nitrogen) flow rates at 35 and 18 arbitrary units, respectively; capillary temperature at 290 ◦ C, the capillary voltage at 45 V, the tube lens voltage at 90 V, the skimmer voltage at 22 V. Data were acquired in full scan mode over a mass range of 80–500 m/z. The instrument operated in positive ion mode with a resolving power of 100,000 FWHM. In case of confirmation of screening positive results, a further set of experiments was performed with in-source collision-induced dissociation (CID) with voltage set at 40 V, acquiring ions from 50 to 500 m/z, with a resolving power of 50,000 FWHM, obtaining the accurate masses of both precursor and fragment ions. Mass calibration was optimized according to the guidelines provided by the instrument supplier. The recommended calibration solution was made of MRFA (l-methionyl-arginyl-phenylalanylalanine acetate) 1 g mL−1 , caffeine 2 g mL−1 and Ultramark® 1621 0.001% dissolved in methanol/water (1:1). For calibration the automatic calibration feature of the Exactive tune software was
146
S. Strano-Rossi et al. / Journal of Pharmaceutical and Biomedical Analysis 106 (2015) 144–152
used. The mass scale required to be calibrated every 2 days over the mass range m/z 50–2000. Lock-mass (diisooctyl phtalate ionic species, 391.2843 m/z) was employed during the analysis of samples in order to compensate any possible mass axis drifts. 2.3.3. Data analysis For qualitative screening, data were processed using a homemade database containing the protonated exact monoisotopic masses list and built on the extracted ion chromatograms (EIC) of the expected [M+H]+ ions of each compound. The database was created by analysing pure standard solutions. Identification of compounds was based on the following criteria: retention time window (±0.2 min); exact mass of monoisotopic ion M+0 (mass accuracy m, defined as (exact mass) − (accurate mass)/(accurate mass) × 106 , less than 2 ppm); comparison of experimental and calculated isotopic pattern (Relative Isotopic Abundances RIA (M+1/M+0) and RIA (M+2/M+0) errors less than 15%). Xcalibur 2.1 software (ThermoFisher Scientific Inc., San Jose, CA, USA) was used to analyze and process all data for quantitative analysis. 2.4. Method validation The performance of the method was evaluated through estimation of the following parameters: specificity and selectivity, limit of detection (LOD), limit of quantitation (LOQ), linearity, repeatability, accuracy, matrix effect, recovery and carryover. Specificity of the method was studied by analyzing ten nutritional supplement samples free of stimulants in order to assess potential interferences from other substances. There were different kinds of nutritional supplements: energy bars, tablets and powders (in some cases vanilla and cocoa flavored) and oily solutions. Selectivity of the method was studied by analysing samples spiked with various classes of drugs such as cocaine, new psychoactive drugs, opiates, antidepressants, benzodiazepines, cannabinoids and steroids. The LOD and the LOQ were experimentally determined using the signal-to-noise approach by analysing serial dilutions (1–5–10–25 ng g−1 ) of fortified samples on five different typologies of supplements. LODs were figured out at the lowest concentration value that provided a S/N > 3 for the extracted ionic trace, while LOQs at the minimum concentration giving a S/N > 10 along with an acceptable precision (%CV < 20%) and accuracy (%E < 20%). Linearity was investigated over a five-point calibration with spiked dietary supplement ranging from the LOQ of each substance (50 ng g−1 ) up to 2000 ng g−1 . Calibration curves were built by linear regression, considering the area ratio between the analyte and the IS, and were prepared in triplicate by adding the proper amount of working mixture to the aliquots of 1 g of nutritional supplement in order to obtain the following concentrations: 50–100–500–1000–2000 ng g−1 . Repeatability and accuracy, expressed as %CV and %E, respectively, were assessed on control samples prepared at three different concentrations, i.e., 50, 500 and 2000 ng g−1 by quintuplicate analyses in three different days. Matrix effects were investigated for each compound by comparing the response (in terms of peak areas) of a mixture of five blank supplements post-fortified with the analyte to those obtained injecting neat solutions prepared at the same concentrations. According to Matuszewski [17] absolute recoveries were determined by comparing the signals of QC samples to those obtained spiking blanks supplements after extraction at the same concentrations. Carryover was evaluated through the injections of two blank samples after each calibration level and evaluating the eventual presence of memory effect.
3. Results The proposed method allowed the simultaneous screening of various stimulants/anorectics and PDE5Is in nutritional supplements, using the same sample preparation procedure and guaranteeing optimal results in terms of sensitivities. The preparation step with sodium hydroxide and the following extraction were fundamental in order to have a cleaner extract and therefore to lower the ion suppression, allowing to obtain satisfactory sensitivities. The screening of compounds was performed in full scan experiment and preliminary identification was achieved by evaluation of retention time, accurate mass (experimental mass accuracy, m, obtained with lockmass, was less than 1 ppm for all ionic species) and experimental and calculated isotopic cluster correspondence. The difference between experimental and calculated M+1/M+0 and M+2/M+0 abundance ratios was within 15% for all substances. The MH+ elemental composition, the exact and accurate masses, obtained m and retention times are reported in Table 1. The methodology was validated through assessment of linearity, specificity, sensibility, precision, accuracy, matrix effect. Table 2 are reported LODs, LOQs and linearity (mean slopes, intercepts and correlation coefficients obtained) for each substance included in the present study. LODs varied from less than 1 ng g−1 , in case of caffeine, clenbuterol, fencamfamine, fenfluramine and pseudoephedrine, to 25 ng g−1 for phentermine, vardenafil and sildenafil. All the studied compounds were quantified at ng g−1 levels. The method showed good linearity in the range from the LOQ of each substance to 2000 ng g−1 of nutritional supplement. In case of supplements containing substances at concentration above the quantitation range, a lesser amount of supplement (1 mg instead of 100 mg) was submitted to the analytical protocol. Also accuracy, intended as mean relative error (%E) and interand intra-day repeatability (CV%), were always lower than 15%. Recovery was acceptable for all substances, ranging from 27% for benfluorex at low concentration to over 50% for fencanfamine in the case of stimulants and from 33%to 44% for PDE5Is. Data on recoveries, interday repeatability and accuracy are shown in Table 3. No interferences from other ingredients with the same accurate masses were encountered at the retention times of the analytes. Fig. 1 shows the layout of a blank nutritional supplement sample spiked with the analytes of interest at a concentration of 100 ng g−1 . The results on the matrix effect evaluation are listed in Table 3 with a value of 100% indicating no matrix effect, 100% indicating signal enhancement. In particular, as it was expected due to the complexity of a multicomponent matrix, a matrix effect was observed for some analytes. The highest signal suppression was observed for benfluorex at 27% and highest signal enhancement was observed for tadalafil at 8%. These values can be considered acceptable. No carryover effect had been observed, apart for ephedrines after the analysis of supplements containing high concentrations of these substances. In these cases three blank samples were injected afterwards, and a complete decontamination of the system was ensured by the complete negativity of the third blank injection. The characteristic fragmentation pattern of the analytes was used to further confirm positive cases. This was obtained by performing a further collision experiment and by identifying the compound of interest, in addition to its retention time, according to the relative abundance of the generated characteristic fragments and their accurate masses with respect to those obtained by a reference standard. Relative ion intensities should not vary more than ±20% for ions with relative intensities >50%, ±25% for ions with relative intensities between 10% and 50% and ±50% for ions with
S. Strano-Rossi et al. / Journal of Pharmaceutical and Biomedical Analysis 106 (2015) 144–152
147
Table 1 MH+ elemental composition, exact and accurate masses, obtained m and retention times of analytes included in the study. Compound
MH+ elemental composition
Exact mass (m/z)
Accurate mass (mean m/z)
m (mean ppm)
Benfluorex Caffeine Clenbuterol Ephedrine/pseudoephedrine Fencamfamine Fenfluramine Mephentermine Phendimetrazine Phenmetrazine Phentermine Sibutramine Sildenafil Tadalafil Vardenafil
C19 H21 F3 NO2 C8 H11 N4 O2 C10 H19 Cl2 N2 O C10 H16 NO C15 H22 N C12 H17 F3 NO C11 H18 N C12 H18 NO C10 H16 NO C10 H16 N C17 H27 ClN C22 H31 N6 O4 S C22 H20 N3 O4 C23 H33 N6 O4 S
352.1519 195.0877 277.0869 166.1226 216.1747 232.1308 164.1434 192.1383 178.1226 150.1277 280.1827 475.2122 390.1448 489.2279
352.1515 195.0874 277.0866 166.1224 216.1744 232.1304 164.1432 192.1381 178.1224 150.1276 280.1822 475.2122 390.1449 489.2280
−1.1 −1.1 −1.0 −1.2 −1.2 −1.6 −1.2 −0.8 −1.2 −0.9 −1.6 0.1 0.1 0.3
Table 2 Limits of detection, limits of quantitation and linearity in the range from 50 to 2000 ng g−1 . Compound
LOD (ng g−1 )
LOQ (ng g−1 )
Slope ± SD
Benfluorex Caffeine Clenbuterol Fencamfamine Fenfluramine Mephentermine Phendimetrazine Phenmetrazine Phentermine Pseudoephedrine Sibutramine Sildenafil Tadalafil Vardenafil
10 1 1 1 1 10 5 5 25 1 10 25 5 25
50 50 50 50 50 50 50 50 50 50 50 50 50 50
0.0488 0.0250 0.2452 0.2210 0.0124 0.0662 0.0332 0.0637 0.0028 0.0192 0.0523 0.0660 0.0219 0.0013
relative intensities < 10% in the unknown and in the reference, and mass accuracies of characteristic ions should not vary more than 5 ppm. This method was applied to 36 nutritional supplements sold in fitness centers and in websites, collected from police officers in Italy. HRMS analysis showed the presence of ephedrine in three samples at a concentration ranging from 370 ng g−1 to 1000 ng g−1 (in one case ephedrine was not declared in the label), ephedrine and pseudoephedrine in one sample at 460 ng g−1 and 540 ng g−1 , respectively. A supplement, that was supposed to contain ephedrine according to the label, was negative. Moreover, caffeine, whose presence was reported in the label of six-supplement only, was detected at concentrations ranging from 225 up to 40500 ng g−1 in eight supplements. In the cases where caffeine
± ± ± ± ± ± ± ± ± ± ± ± ± ±
Intercept ± SD
0.0043 0.0009 0.0095 0.0151 0.0003 0.0053 0.0040 0.0055 0.0004 0.0026 0.0092 0.0110 0.0108 0.0004
0.1722 0.0367 −2.3048 −1.7875 −0.7341 −0.0631 2.0147 0.8948 0.4399 −0.0597 0.1722 −0.1139 −0.5413 −0.0155
± ± ± ± ± ± ± ± ± ± ± ± ± ±
r2 ± SD
0.6062 0.0041 9.0434 3.6341 0.7991 0.1601 1.2860 1.5011 0.3337 0.2609 0.4224 0.1306 0.5864 0.0159
0.9971 0.9905 0.9924 0.9926 0.9956 0.9955 0.9951 0.9943 0.9944 0.9959 0.9917 0.9969 0.9957 0.9987
± ± ± ± ± ± ± ± ± ± ± ± ± ±
0.0013 0.0009 0.0007 0.0016 0.0029 0.0010 0.0033 0.0005 0.0014 0.0030 0.0004 0.0009 0.0046 0.0004
concentration exceeded the upper level of the calibration curve, the quantification was obtained by analysing only one milligram of each supplement. In four batches of tablets called “Kamagra” sildenafil was identified and, after proper dilution, quantified at a mean concentration of 200 mg g−1 , corresponding to nearly 100 mg per tablet. The extracted ionic chromatograms and mass spectra of a nutritional supplement sample containing ephedrine at 460 ng g−1 , pseudoephedrine at 540 ng g−1 and caffeine at 17800 ng g−1 are shown in Fig. 2. In Fig. 3a it is represented the isotopic cluster corresponding to the raw formula of sildenafil (C22 H31 N6 O4 S), relative to a Kamagra tablet. Fig. 3b shows the comparison of fragments obtained by CID experiments from Kamagra tablets and a sildenafil standard.
Table 3 Accuracy (E%), precision (expressed as CV%), matrix effect (ME%) and recovery (R%) data for selected analytes. Compound
Benfluorex Caffeine Clenbuterol Fencamfamine Fenfluramine Mephentermine Phendimetrazine Phenmetrazine Phentermine Pseudoephedrine Sibutramine Sildenafil Tadalafil Vardenafil
50 ng g−1
500 ng g−1
2000 ng g−1
E%
CV%
ME%
R%
E%
CV%
ME%
R%
E%
CV%
ME%
R%
8 −3 11 9 3 −12 1 4 7 3 −9 3 3 1
5 8 12 12 7 7 8 10 12 3 5 1 2 2
73 86 83 96 95 97 80 75 90 74 91 93 99 98
24 42 36 44 33 33 33 27 36 29 32 34 36 44
6 13 7 3 −8 8 2 −2 1 −2 −7 −3 −1 −3
8 8 6 8 6 3 8 8 15 11 4 5 5 2
88 93 93 99 101 95 101 107 105 107 98 94 101 107
45 36 30 54 41 33 43 35 37 28 35 41 33 33
4 −2 −1 1 3 −2 3 1 −2 2 −3 1 3 2
3 5 2 3 2 1 3 4 5 4 1 3 1 0.4
88 93 90 96 95 92 100 95 97 96 103 91 108 94
36 39 49 42 45 34 35 29 42 29 42 41 44 39
148
S. Strano-Rossi et al. / Journal of Pharmaceutical and Biomedical Analysis 106 (2015) 144–152 12.22
tadalafil
100
Tadalafil
50
6.45
sibutramine
0 100 50 0.97
benfluorex
9.30
3.72
sildenafil vardenafil
6.18
fencamfa.. fenfluram.. Clenbuterol
NL: 2.92E5 m/z= 475.2074-475.2170 MS 100ng_g_smart_sto p_2_stimolanti
10.29
Sildenafil
50
2.95 3.31
4.41 4.94
7.40
5.84
11.24
NL: 3.73E4 m/z= 489.2230-489.2328 MS 100ng_g_smart_sto p_2_stimolanti
9.65
Vardenafil
50
9.80
NL: 1.82E6 m/z= 216.1725-216.1769 MS 100ng_g_smart_sto p_2_STIMOLANTI
8.81
Fencamfamine 50 8.95 8.15
9.21 10.26
11.12
NL: 6.61E5 m/z= 232.1285-232.1331 MS 100ng_g_smart_sto p_2_STIMOLANTI
8.80
Fenfluramine
50
9.28
0 100
7.39
Clenbuterol
50 0.63
0 100 Caffeine
11.33
8.62
0 100
5.13
6.42 6.66
8.28
9.04 9.36 9.88 10.34
11.58
12.30 12.85
Caffeine
50
0.14 0.40 1.26 1.95 2.29 2.68 3.26
3.90
5.43 5.98
4.67
6.76
10.12 10.21 10.28 11.32
7.38 7.74 8.40
13.54
NL: 4.59E5 m/z= 277.0841-277.0897 MS 100ng_g_smart_sto p_2_stimolanti
NL: 1.15E5 m/z= 195.0857-195.0897 MS 100ng_g_smart_sto 13.77 p_2_stimolanti
7.22
0 100 Phenterm..
NL: 4.56E5 m/z= 352.1484-352.1554 MS 100ng_g_smart_sto p_2_STIMOLANTI
Benfluorex
50
0 100
12.59
NL: 4.75E4 m/z= 150.1262-150.1292 MS 100ng_g_smart_sto 6.35 p_2_stimolanti 7.31 8.43 8.78 9.61 9.96 10.23 11.27 13.66 0.68 1.10 1.66 2.05 2.86 3.21 3.69 4.32 4.92 5.43 5.85 12.06 12.87 NL: 1.53E5 6.59 m/z= 164.1418-164.1450 MS 100ng_g_smart_sto p_2_stimolanti 6.91 7.49 7.88 8.60 9.13 1.39 3.09 12.02 12.83 13.64 9.89 10.61 11.08 NL: 1.25E7 6.31 m/z= 192.1364-192.1402 MS 100ng_g_smart_sto p_2_stimolanti 7.04 7.63 7.93 8.44 9.58 9.82 10.33 11.05 11.60 12.25 12.78 13.95 0.70 1.25 1.69 2.40 2.81 3.70 3.97 4.87 5.66 NL: 1.60E7 6.24 m/z= 178.1208-178.1244 MS 100ng_g_smart_sto p_2_stimolanti 7.01 7.63 7.97 8.52 9.14 9.59 10.00 10.54 11.13 12.16 12.86 0.71 1.14 2.01 2.50 3.41 3.83 4.37 4.75 5.53 13.95 NL: 1.28E5 5.24 0.68 m/z= 166.1209-166.1243 MS 100_ng_g_smart_st 5.42 op_2_stimolanti 1.07 5.66 6.22 7.14 1.25 2.31 7.91 8.22 0.23 3.29 4.05 4.62 4.84 9.30 9.79 10.09 10.96 11.79 12.79 13.42 7.04
Phentermine
50 0 100
Mephent..
12.09 12.68
10.06 11.12
0 100
Phendimetr..
NL: 7.52E5 m/z= 280.1799-280.1855 MS 100ng_g_smart_sto p_2_STIMOLANTI
Sibutramine
0 100
Phenmetraz..
13.59
11.61
0 100
Pseudoeph..
12.67
10.11
NL: 3.85E5 m/z= 390.1409-390.1487 MS 100ng_g_smart_sto p_2_stimolanti
Mephentermine
50 0 100
Phendimetrazine
50 0 100
Phenmetrazine
50 0 100
Pseudoephedrine
50 0 0
1
2
3
4
5
6
7 Time (min)
8
9
10
11
12
13
14
Fig. 1. HR ion chromatograms, based on accurate mass of each analyte, of a nutritional supplement sample spiked with the analytes of interest at a concentration of 100 ng g−1 .
S. Strano-Rossi et al. / Journal of Pharmaceutical and Biomedical Analysis 106 (2015) 144–152
149 NL: 2.58E8 m/z= 195.0857195.0897 MS 130624_ca mp3
7.12
100 90
Caffeine
80 Relative Abundance
70 60 50 40 30 20
7.47 7.92 8.37 8.74 8.99 9.48
10 0.06 0.49
0 0
1
2
3
4
5
6
7 Time (min)
8
9
10
11
12.67 13.17 13.52 13.92 12
MH+
80
experimental isotopic paern
60 40 20
13
14
NL: 1.87E8 130624_camp3#528-536 RT: 7.09-7.20 AV: 9 SB: 26 6.82-7.00 , 7.41-7.56 T: FTMS {1,1} + p ESI Full ms [80.00-500.00]
195.0875
100 Relative Abundance
10.32 10.72 11.19
5.65 5.99 6.43 6.71
196.0908 195.0132 195.1619 195.3110 195.0877
0 100
195.5315 195.6792
195.9329
197.0915 NL: 2.11E4
MH+
80
calculated isotopic paern
C 8 H 10 N 4 O 2 +H: C 8 H 11 N 4 O 2 p (gss, s /p:40) Chrg 1 R: 100000 Res .Pwr . @FWHM
60 40 20
196.0911 197.0919
0 195.0
195.2
195.4
195.6
195.8
196.0
196.2
196.4
196.6
196.8
197.0
197.2
m/z
90
5.06
Ephedrine Pseudoephedrine
80 Relative Abundance
NL: 1.10E5 m/z= 166.1209166.1243 MS 130624_ca mp3
5.22
100
70 60 50 40 30 20 10 0 .2 5
1.02
1.54
2.19
2.92 3.27
3 .9 2
5.46 5.63 6.13
4 .6 5
7.14 7.56 7.93
0 0
2
3
4
5
6
7 Time (min)
8
8.10
9.08 9.47 9.82 10.43 10.97 11.36 12.03 12.58 9
10
11
12
80
MH+ experimental isotopic paern
60 40 20
167.1256 166.8371 166.9987
0 100
13.32 13
14
NL: 4.64E4 130624_camp3#583-594 RT: 5.01-5.11 AV: 12 SB: 10 4.92-5.00 , 5.29-5.37 T: FTMS {1,1} + p ESI Full lock ms [50.00-550.00]
166.1224
100 Relative Abundance
1
167.5476
168.1277 NL: 2.09E4 C 10 H 15 NO +H: C 10 H 16 N 1 O 1 p (gss, s /p:40) Chrg 1 R: 100000 Res .Pwr . @FWHM
80
MH+ calculated isotopic paern
60 40 20
167.1260 168.1293
0 166.0
166.2
166.4
166.6
166.8
167.0
167.2 m/z
167.4
167.6
167.8
168.0
168.2
168.4
Fig. 2. HR ion chromatograms, based on accurate mass of each analyte, and experimental and calculated MH+ isotopic patterns of a nutritional supplement sample containing ephedrine at 460 ng g−1 , pseudoephedrine at 540 ng g−1 and caffeine at 17,800 ng g−1 .
150
S. Strano-Rossi et al. / Journal of Pharmaceutical and Biomedical Analysis 106 (2015) 144–152
Fig. 3. (a) [M + H]+ isotopic cluster of the component of a Kamagra tablet, corresponding to the raw formula of sildenafil (C22 H31 N6 O4 S); and (b) fragments obtained in CID experiments at 40 V for kamagra tablets (upper box) and sildenafil standard (lower box).
4. Discussion The identification of stimulants and PDE5Is in nutritional supplements is a matter of concern for the public health. In some cases they can be even not declared in the label, and this poses a risk to the
health, especially on subjects with pathologies that can be worsen by the above mentioned drugs, or on professional athletes, that can be exposed to an adverse analytical finding during antidoping tests. For instance, the presence of benfluorex, although not expressly included in the WADA list of prohibited substances, can lead to an
S. Strano-Rossi et al. / Journal of Pharmaceutical and Biomedical Analysis 106 (2015) 144–152
adverse analytical finding for norfenfluramine, its main metabolite. Other studied classes of substances, such as stimulants and anorectics, besides being included in the WADA prohibited list, may be dangerous for the health and for this reason some of them have been withdrawn from the market in many Countries [18–21]. Even when they are declared, such as in the case of kamagra containing sildenafil, these kinds of drugs, when prepared by unknown laboratories and sold on-line, are not subjected to the national and international controls, that are compulsory for pharmaceutical products, and could be potentially dangerous. A method for the identification of these drugs is therefore needed. The proposed approach provides the use of a sodium hydroxide purification step prior to the extraction using a mixture of solvents. With this preparation, a clean extract, even in the case of oily supplements or energy bars, can be obtained, and all the classes of analytes investigated are satisfactorily extracted. In one paper reporting the analysis of clenbuterol in tablets is described a similar approach with the use of diluted KOH for dissolving the sample and extraction with tertbutyl-methyl ether [3]. Other studies perform a simple dissolving of supplements in methanol [6,7,12] or acetonitrile/water [11] and a subsequent filtration, obtaining good recoveries for the analysis of anorectic compounds and PDE5Is; these approaches, however, cannot be applied in case of complex matrices such as bars or oily preparations. The optimized chromatographic gradient allows a good separation of analytes, also for similar compounds. For example, in the case of ephedrines, it was possible the discrimination of isomers (ephedrine/pseudoephedrine), as can be seen in Fig. 2. Moreover, the full scan acquisition enables the identification of other analytes not included in the design of method development. Indeed, it was possible to identify traces of norephedrine/norpseudoephedrine in the samples containing ephedrines. Both isomers were detected in the sample with ephedrine and pseudoephedrine, whilst only traces of norephedrine and methylephedrine were found in one of the samples showing higher amounts of ephedrine. On the other end, the mass spectrometric analysis performed at high mass resolution enhances the specificity of the method for the seeked compounds and enables the evaluation of the fine molecular structures, allowing the reliable identification of the elemental composition of analytes. In case of kamagra, for example, the use of a resolution of 100,000 (FWHM at m/z 200) allows to evince the presence of a sulfur atom, as it is possible to observe a splitting of the [M+2] isotopes (see Fig. 3a). The resolved signal at m/z 477.2080 in fact corresponds to the contribute of the isotope 34 S, which gives a theoretical relative mass with +1.996 with respect to the M+0 monoisotopic ion (m/z 475.2122), while the second peak at m/z 477.2181 corresponds to the contribute of 13 C2 .
5. Conclusions This study reports the application of a method for the analysis of dietary supplements for critical compounds potentially dangerous for the public health, especially when they are not reported in the labels. By using the benefits of the versatility of liquid chromatography coupled to the high resolution mass spectrometry, a screening method was developed and validated for the analysis of stimulants, anorectics and PDE5Is in nutritional supplements. Cheap and efficient sample preparation step enabled an effective purification of the supplements, and HRMS enables high selectivity for the monitoring of compounds that share the same nominal mass but differ in their elemental composition and therefore have a different exact mass, and therefore allows to distinguish between analytes of interest and isobaric contaminants coming from a
151
complex matrix such as nutritional supplements. Moreover, high resolution mass spectrometry allows the accurate identification of analytes, through the exact masses of target compounds and the associated isotopic patterns. The great potentiality of high resolution mass spectrometry technique is also related to the possibility of reprocessing previously acquired data files when a further substance has to be investigated. The application of the method to the analysis of 36 dietary supplements, sold in fitness centers and in unauthorized online markets, allowed the detection of substances not declared in the products labels, such as ephedrine and caffeine. This highlights the importance of the analysis of dietary supplements as they are highly consumed and sometimes misused by general public unaware of the potential health risk. Acknowledgements This work was carried out in the framework of the project “Smart Stop” – Control on the sites selling products dangerous for the public health, financed by the Antidrug Policies Department, Presidency of Italian Council of Ministers. References [1] Directive 2002/46/EC, http://ec.europa.eu/food/food/labellingnutrition/ supplements/index en.htm (accessed 20.04.14). [2] Directive 2010/84/EU, http://ec.europa.eu/health/files/eudralex/vol-1/dir 2010 84/dir 2010 84 en.pdf (accessed 20.04.14). [3] M.K. Parr, K. Koehler, H. Geyer, S. Guddat, W. Schanzer, Clenbuterol marketed as dietary supplement, Biomed. Chromatogr. 22 (2008) 298–300. [4] J. Jung, M. Hermanns-Clausen, W. Weinmann, Anorectic sibutramine detected in a Chinese herbal drug for weight loss, Forensic Sci. Int. 161 (2006) 221–222. [5] C. Vidal, S. Quandte, Identification of a sibutramine-metabolite in patient urine after intake of a “pure herbal” Chinese slimming product, Ther. Drug Monit. 28 (2006) 690–692. [6] P. Zou, S.S. Oh, K.H. Kiang, M.Y. Low, B.C. Bloodworth, Detection of sibutramine, its two metabolites and one analogue in a herbal product for weight loss by liquid chromatography triple quadrupole mass spectrometry and time-of-flight mass spectrometry, Rapid Commun. Mass Spectrom. 21 (2007) 614–618. [7] H.J. Kim, J.H. Lee, H.J. Park, S.H. Cho, S. Cho, W.S. Kim, Monitoring of 29 weight loss compounds in foods and dietary supplements by LC–MS/MS, Food Addit. Contam. A: Chem. Anal. Control Expo. Risk Assess. 31 (2014) 777–783. [8] F.T. Delbeke, Nutritional supplements and doping, in: C. Peters, T. Schulz, H. Michna (Eds.), Biomedical Side Effects of Doping, Cologne, 2001, pp. 155–161. [9] O. de Hon, B. Coumans, The continuing story of nutritional supplements and doping infractions, Br. J. Sports Med. 41 (2007) 800–805, discussion 805. [10] C.N. Man, N.M. Nor, R. Lajis, G.L. Harn, Identification of sildenafil, tadalafil and vardenafil by gas chromatography–mass spectrometry on short capillary column, J. Chromatogr. A 1216 (2009) 8426–8430. [11] F. Shi, C. Guo, L. Gong, J. Li, P. Dong, J. Zhang, P. Cui, S. Jiang, Y. Zhao, S. Zeng, Application of a high resolution benchtop quadrupole–Orbitrap mass spectrometry for the rapid screening, confirmation and quantification of illegal adulterated phosphodiesterase-5 inhibitors in herbal medicines and dietary supplements, J. Chromatogr. A. 1344 (2014) 91–98. [12] H.M. Lee, B.J. Lee, A novel approach to simultaneous screening and confirmation of regulated pharmaceutical compounds in dietary supplements by LC/MS/MS with an information-dependent acquisition method, Food Addit. Contam. A: Chem. Anal. Control Expo. Risk Assess. 28 (2011) 396–407. [13] C.M. Gryniewicz, J.C. Reepmeyer, J.F. Kauffman, L.F. Buhse, Detection of undeclared erectile dysfunction drugs and analogues in dietary supplements by ion mobility spectrometry, J. Pharm. Biomed. Anal. 49 (2009) 601–606. [14] B.J. Venhuis, D. de Kaste, Towards a decade of detecting new analogues of sildenafil, tadalafil and vardenafil in food supplements: a history, analytical aspects and health risks, J. Pharm. Biomed. Anal. 69 (2012) 196–208. [15] Wada, List of Prohibited Substances and Methods, 2014, http://www.wadaama.org/Documents/World Anti-Doping Program/WADP-Prohibited-list/ 2014/WADA-prohibited-list-2014-EN.pdf (accessed 20.04.14). [16] S. Martello, M. Felli, M. Chiarotti, Survey of nutritional supplements for selected illegal anabolic steroids and ephedrine using LC-MS/MS and GC–MS methods, respectively, Food Addit. Contam. 24 (2007) 258–265. [17] B.K. Matuszewski, M.L. Constanzer, C.M. Chavez-Eng, Strategies for the assessment of matrix effect in quantitative bioanalytical methods based on HPLC–MS/MS, Anal. Chem. 75 (2003) 3019–3030.
152
S. Strano-Rossi et al. / Journal of Pharmaceutical and Biomedical Analysis 106 (2015) 144–152
[18] A.J. Scheen, Cardiovascular risk-benefit profile of sibutramine, Am. J. Cardiovasc. Drugs 10 (2010) 321–334. [19] C. Tribouilloy, S. Marechaux, Y. Jobic, A. Jeu, S. Ederhy, E. Donal, P. Reant, E. Arnalsteen, J. Boulanger, T. Garban, P.V. Ennezat, M. Andrejak, D. Rusinaru, Frequency of drug-induced valvular heart disease in patients previously exposed to benfluorex: a multicentre prospective study, Eur. Heart J. 34 (2013) 3580–3587.
[20] M. Andrejak, C. Tribouilloy, Drug-induced valvular heart disease: an update, Arch. Cardiovasc. Dis. 106 (2013) 333–339. [21] L. Abenhaim, Y. Moride, F. Brenot, S. Rich, J. Benichou, X. Kurz, T. Higenbottam, C. Oakley, E. Wouters, M. Aubier, G. Simonneau, B. Begaud, Appetite-suppressant drugs and the risk of primary pulmonary hypertension, N. Engl. J. Med. 335 (1996) 609–616.
Contents lists available at ScienceDirect
Journal of Pharmaceutical and Biomedical Analysis journal homepage: www.elsevier.com/locate/jpba
Liquid chromatography–high resolution mass spectrometry (LC–HRMS) determination of stimulants, anorectic drugs and phosphodiesterase 5 inhibitors (PDE5I) in food supplements Sabina Strano-Rossi a,∗ , Sara Odoardi a , Erika Castrignanò a , Giovanni Serpelloni b , Marcello Chiarotti a a b
Institute of Public Health, Forensic Toxicology Laboratory, Università Cattolica del S. Cuore, Rome, Italy Antidrug Policies Department, Presidency of Council of Ministers, Rome, Italy
a r t i c l e
i n f o
Article history: Received 5 May 2014 Received in revised form 6 June 2014 Accepted 7 June 2014 Available online 14 June 2014 Keywords: LC–HRMS Nutritional supplements Stimulants PdE5Is Anorectics
a b s t r a c t The paper describes a liquid chromatography/high resolution mass spectrometry LC/HRMS method for the simultaneous identification and quantification of stimulants (ephedrines, caffeine, anorectic drugs such as phentermine, phendimetrazine, phenmetrazine, fenfluramine, benfluorex, mephentermine, fencanfamine, sibutramine) and PDE5I (sildenafil, vardenafil and tadalafil) in food supplements using a benchtop Orbitrap mass spectrometer. The mass detector, with a nominal resolving power of 100,000 (FWHM at m/z 200), operated in full scan mode in ESI positive ionization mode. Analytes were identified by retention times, accurate masses and correspondence of experimental and calculated isotopic patterns. The limits of detection (LOD) obtained varied from 1 to 25 ng g−1 and limits of quantification (LOQ) were 50 ng g−1 for all compounds. The method was linear for all the analytes in the ranges from 50 to 2000 ng g−1 , giving correlation coefficients > 0.99. Accuracy (intended as %E) and repeatability (% CV) were always lower than 15%. The method was applied to the analysis of 36 dietary supplements, revealing the presence of ephedrine and/or pseudoephedrine in four of them, caffeine in eight of them and sildenafil in four of them. In one case, ephedrine was not reported on the label of the dietary supplement, as well as for caffeine in other two cases. A further confirmation of the analytes identity in positive samples was obtained through in-source fragmentation and comparison of the obtained fragments and their relative abundances with those from certified standards. As the acquisition mode is full scan, it would be also possible to re-process a previously acquired datafile for the investigation of untargeted analytes. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Dietary supplements are concentrated sources of nutrients that are theoretically present in a normal and balanced diet, such as minerals, vitamins, amino acids, trace elements and other components whose purpose is to supplement the normal diet. They are sold in form of tablets, capsules or in powder for an oral consumption. All the layers of population are prone to intake nutritional supplements and herbal remedies, as these preparations are generally considered safer than synthetic drugs and claim to benefit consumers for treating various health conditions and promoting
∗ Corresponding author at: Institute of Public Health, Forensic Toxicology Laboratory, Università Cattolica del S. Cuore, l.go F. Vito, 1, Rome, Italy. Tel.: +39 0630154249; fax: +39 063051168. E-mail addresses: [email protected], [email protected] (S. Strano-Rossi). http://dx.doi.org/10.1016/j.jpba.2014.06.011 0731-7085/© 2014 Elsevier B.V. All rights reserved.
general wellbeing. The widespread diffusion of Internet sites and gyms selling supplements render difficult the application of international and national directives regarding to the labeling and the controls on food supplements and on pharmaceutical products [1,2]. This poses a risk to the public health and it can be considered a legal offense in many Countries. Furthermore, supplements safety is of major concern, since chemicals can be added illegally to dietary supplements in order to enhance the activities claimed (anorectic, stimulating, erectile function stimulating, diuretic). During recent years it has been reported that products marketed as dietary supplements could contain non-labeled substances, like clenbuterol [3], the anorectic drugs sibutramine [4–7], phentermine and fenfluramine [8], methylenedioxymetamphetamine (MDMA) [9], phosphodiesterase 5 inhibitors (PDE5I) like sildenafil, vardenafil, tadalafil [10–13] and analogs [13,14]. Moreover, all of the above-mentioned substances, with the exception of PDE5Is, are included in the World Anti-Doping Agency (WADA) list of prohibited substances, and their intake, also if
S. Strano-Rossi et al. / Journal of Pharmaceutical and Biomedical Analysis 106 (2015) 144–152
involuntary or accidental, can result in a doping offense for the athlete, giving rise to a positive doping test [15]. In fact not only amateur sports people but also professional athletes are tempted to take dietary supplements as they believe assuming natural and safer products in order to modify their body’s appearance, to promote fat burning and to improve their performances by a stimulating effect. Besides the risk of violating doping rules, both athletes and, more in general, all population are exposed to the high health risk involved in the intake of products that may contain unlabeled drugs with potentially dangerous unwanted effects. The analysis of food supplements, to exclude the presence of drugs, is therefore of fundamental importance for the evaluation of products’ safety, especially when they are sold by gyms as well as through not-conventional channels such as internet websites and “spice shops”. The aim of this paper is to provide a valid method for the simultaneous screening of stimulants and PDE5I in nutritional supplements, performed by Liquid Chromatography (LC) High Resolution Mass Spectrometry (HRMS) using an Exactive benchtop Orbitrap mass spectrometer, after a simple extraction of active ingredients and possible contaminants from food supplements. The main advantage of the use of Orbitrap mass spectrometry for the analysis of complex matrices is the possibility of obtaining a mass resolution till 100,000 (full width at half maximum FWHM) at m/z 200, with a mass accuracy of 2 mDa (as stated by the manufacturer), achieved by using for the internal calibration a lockmass such as the diisooctyl phtalate ionic species, normally present in a laboratory environment. This leads to a high specificity of the methodology for the target compounds, useful for the identification of contaminants in complex matrices like food supplements. This technology, in fact, achieving a high mass resolution and hence a high accuracy of mass measurement, enhances the possibility to distinguish the target analyte from nearly isobaric compounds that, if co-eluting, can interfere in the analysis, even after a prior sample purification step. Possible interferences can therefore be resolved by the use of HRMS. Another main advantage of this technique is that the analyses are performed in full scan mode. This enables a further investigation of a wider range of compounds, even if they were not included at the early stage of the method development, by means of a simple re-elaboration of the previously acquired datafile. 2. Materials and methods 2.1. Chemicals and reagents Formic acid, sodium hydroxide, ultrapure water and methanol were purchased from Sigma–Aldrich (Milano, Italy). All solvents used were LC–MS grade. Clenbuterol was from Sigma–Aldrich (Saint Louis, MO, USA). Caffeine, fenfluramine, phendimetrazine, phenmetrazine, phentermine, mephentermine, ephedrine, pseudoephedrine, norpseudoephedrine, norephedrine, sibutramine, benfluorex and methamphetamine-D5 were from LGC Standards (Milano, Italy); fencamfamine was from Salars (Como, Italy). Methanolic stock solutions (1 mg mL−1 ) of all substances were used for the preparation of the working solution mixture, containing the analytes at a concentration of 10 g mL−1 in methanol. It was stored in freezer at −20 ◦ C. 2.2. Sample preparation Nutritional supplements in powder, oily capsules and tablets seized in fitness leisure centers or sold in the internet market were analyzed in order to check the eventual presence of unlabeled substances. As declared on the labels, they contained proteins, vitamins
145
and fruit, chocolate or vanilla flavors, caffeine in six supplements (labeled as guaranà in four of them) and ephedrines (labeled as ephedrine or ma huang, in four of them), or, in the case of a product called “Kamagra”, sildenafil. Aliquots of these nutritional supplements were collected in glass tubes and kept refrigerated at 4 ◦ C in the dark in order to avoid photo-degradation. 100 mg of the sample were weighted, the internal standard (IS) methamphetamine-D5 was added to the samples at 50 ng g−1 and five milliliters of methanol were added. The specimen was then vortexed for 20 s and centrifuged for 10 min at 4000 rpm. The methanolic phase was evaporated to dryness under nitrogen flow at 40 ◦ C and reconstituted in 2 mL of a solution of sodium hydroxide 1 M. A liquid/liquid extraction was performed through the addiction of five milliliters of a mixture of pentane/ethyl ether 9:1, under agitation for an hour [16]. After centrifugation, the organic phase was evaporated to dryness under a nitrogen stream at room temperature and reconstituted with methanol/water/formic acid 6:4:0.03 (100 L) in a glass vial. 10 L were then directly injected in LC–HRMS system. 2.3. LC–HRMS 2.3.1. Equipment The LC–HRMS system was composed of a Thermo ULTIMATE 3000 system equipped with an analytical column Thermo Acclaim RSLC 120 C18 (2.1 mm × 100 mm, 2.2 m particle size), coupled to a Thermo single-stage Orbitrap (Exactive) MS system. The whole equipment and the column were provided by Thermo Fisher Scientific (Milan, Italy).
2.3.2. LC–HRMS conditions The analytical column was maintained at 40 ◦ C, while the sample manager at 15 ◦ C. The following mobile phases were used: (A) ultrapure water/methanol/formic acid 9/1/0.1, (B) methanol acidified with 0.1% formic acid. Separation was achieved using the following mobile phase gradient: 100% A for 1 min, from 0% to 10% of phase B in 0.1 min, linear gradient to 15% B in 4 min, linear gradient to 50% B in 1.8 min, to 70% B in 1.7, to 80% B in 1.1 min, to 100% in 1 min held for 3.5 min. The injection volume was 10 L and the flow rate was set at 0.4 mL min−1 . The chromatographic run time was of 14.5 min per analysis. Ionization parameters were optimized with direct infusion of each substance, including IS methamphetamine-D5, by an external syringe at 10 L min−1 . The chosen tune for the HESI source had the following settings: source current at 5 A, sheath gas and auxiliary gas (either nitrogen) flow rates at 35 and 18 arbitrary units, respectively; capillary temperature at 290 ◦ C, the capillary voltage at 45 V, the tube lens voltage at 90 V, the skimmer voltage at 22 V. Data were acquired in full scan mode over a mass range of 80–500 m/z. The instrument operated in positive ion mode with a resolving power of 100,000 FWHM. In case of confirmation of screening positive results, a further set of experiments was performed with in-source collision-induced dissociation (CID) with voltage set at 40 V, acquiring ions from 50 to 500 m/z, with a resolving power of 50,000 FWHM, obtaining the accurate masses of both precursor and fragment ions. Mass calibration was optimized according to the guidelines provided by the instrument supplier. The recommended calibration solution was made of MRFA (l-methionyl-arginyl-phenylalanylalanine acetate) 1 g mL−1 , caffeine 2 g mL−1 and Ultramark® 1621 0.001% dissolved in methanol/water (1:1). For calibration the automatic calibration feature of the Exactive tune software was
146
S. Strano-Rossi et al. / Journal of Pharmaceutical and Biomedical Analysis 106 (2015) 144–152
used. The mass scale required to be calibrated every 2 days over the mass range m/z 50–2000. Lock-mass (diisooctyl phtalate ionic species, 391.2843 m/z) was employed during the analysis of samples in order to compensate any possible mass axis drifts. 2.3.3. Data analysis For qualitative screening, data were processed using a homemade database containing the protonated exact monoisotopic masses list and built on the extracted ion chromatograms (EIC) of the expected [M+H]+ ions of each compound. The database was created by analysing pure standard solutions. Identification of compounds was based on the following criteria: retention time window (±0.2 min); exact mass of monoisotopic ion M+0 (mass accuracy m, defined as (exact mass) − (accurate mass)/(accurate mass) × 106 , less than 2 ppm); comparison of experimental and calculated isotopic pattern (Relative Isotopic Abundances RIA (M+1/M+0) and RIA (M+2/M+0) errors less than 15%). Xcalibur 2.1 software (ThermoFisher Scientific Inc., San Jose, CA, USA) was used to analyze and process all data for quantitative analysis. 2.4. Method validation The performance of the method was evaluated through estimation of the following parameters: specificity and selectivity, limit of detection (LOD), limit of quantitation (LOQ), linearity, repeatability, accuracy, matrix effect, recovery and carryover. Specificity of the method was studied by analyzing ten nutritional supplement samples free of stimulants in order to assess potential interferences from other substances. There were different kinds of nutritional supplements: energy bars, tablets and powders (in some cases vanilla and cocoa flavored) and oily solutions. Selectivity of the method was studied by analysing samples spiked with various classes of drugs such as cocaine, new psychoactive drugs, opiates, antidepressants, benzodiazepines, cannabinoids and steroids. The LOD and the LOQ were experimentally determined using the signal-to-noise approach by analysing serial dilutions (1–5–10–25 ng g−1 ) of fortified samples on five different typologies of supplements. LODs were figured out at the lowest concentration value that provided a S/N > 3 for the extracted ionic trace, while LOQs at the minimum concentration giving a S/N > 10 along with an acceptable precision (%CV < 20%) and accuracy (%E < 20%). Linearity was investigated over a five-point calibration with spiked dietary supplement ranging from the LOQ of each substance (50 ng g−1 ) up to 2000 ng g−1 . Calibration curves were built by linear regression, considering the area ratio between the analyte and the IS, and were prepared in triplicate by adding the proper amount of working mixture to the aliquots of 1 g of nutritional supplement in order to obtain the following concentrations: 50–100–500–1000–2000 ng g−1 . Repeatability and accuracy, expressed as %CV and %E, respectively, were assessed on control samples prepared at three different concentrations, i.e., 50, 500 and 2000 ng g−1 by quintuplicate analyses in three different days. Matrix effects were investigated for each compound by comparing the response (in terms of peak areas) of a mixture of five blank supplements post-fortified with the analyte to those obtained injecting neat solutions prepared at the same concentrations. According to Matuszewski [17] absolute recoveries were determined by comparing the signals of QC samples to those obtained spiking blanks supplements after extraction at the same concentrations. Carryover was evaluated through the injections of two blank samples after each calibration level and evaluating the eventual presence of memory effect.
3. Results The proposed method allowed the simultaneous screening of various stimulants/anorectics and PDE5Is in nutritional supplements, using the same sample preparation procedure and guaranteeing optimal results in terms of sensitivities. The preparation step with sodium hydroxide and the following extraction were fundamental in order to have a cleaner extract and therefore to lower the ion suppression, allowing to obtain satisfactory sensitivities. The screening of compounds was performed in full scan experiment and preliminary identification was achieved by evaluation of retention time, accurate mass (experimental mass accuracy, m, obtained with lockmass, was less than 1 ppm for all ionic species) and experimental and calculated isotopic cluster correspondence. The difference between experimental and calculated M+1/M+0 and M+2/M+0 abundance ratios was within 15% for all substances. The MH+ elemental composition, the exact and accurate masses, obtained m and retention times are reported in Table 1. The methodology was validated through assessment of linearity, specificity, sensibility, precision, accuracy, matrix effect. Table 2 are reported LODs, LOQs and linearity (mean slopes, intercepts and correlation coefficients obtained) for each substance included in the present study. LODs varied from less than 1 ng g−1 , in case of caffeine, clenbuterol, fencamfamine, fenfluramine and pseudoephedrine, to 25 ng g−1 for phentermine, vardenafil and sildenafil. All the studied compounds were quantified at ng g−1 levels. The method showed good linearity in the range from the LOQ of each substance to 2000 ng g−1 of nutritional supplement. In case of supplements containing substances at concentration above the quantitation range, a lesser amount of supplement (1 mg instead of 100 mg) was submitted to the analytical protocol. Also accuracy, intended as mean relative error (%E) and interand intra-day repeatability (CV%), were always lower than 15%. Recovery was acceptable for all substances, ranging from 27% for benfluorex at low concentration to over 50% for fencanfamine in the case of stimulants and from 33%to 44% for PDE5Is. Data on recoveries, interday repeatability and accuracy are shown in Table 3. No interferences from other ingredients with the same accurate masses were encountered at the retention times of the analytes. Fig. 1 shows the layout of a blank nutritional supplement sample spiked with the analytes of interest at a concentration of 100 ng g−1 . The results on the matrix effect evaluation are listed in Table 3 with a value of 100% indicating no matrix effect, 100% indicating signal enhancement. In particular, as it was expected due to the complexity of a multicomponent matrix, a matrix effect was observed for some analytes. The highest signal suppression was observed for benfluorex at 27% and highest signal enhancement was observed for tadalafil at 8%. These values can be considered acceptable. No carryover effect had been observed, apart for ephedrines after the analysis of supplements containing high concentrations of these substances. In these cases three blank samples were injected afterwards, and a complete decontamination of the system was ensured by the complete negativity of the third blank injection. The characteristic fragmentation pattern of the analytes was used to further confirm positive cases. This was obtained by performing a further collision experiment and by identifying the compound of interest, in addition to its retention time, according to the relative abundance of the generated characteristic fragments and their accurate masses with respect to those obtained by a reference standard. Relative ion intensities should not vary more than ±20% for ions with relative intensities >50%, ±25% for ions with relative intensities between 10% and 50% and ±50% for ions with
S. Strano-Rossi et al. / Journal of Pharmaceutical and Biomedical Analysis 106 (2015) 144–152
147
Table 1 MH+ elemental composition, exact and accurate masses, obtained m and retention times of analytes included in the study. Compound
MH+ elemental composition
Exact mass (m/z)
Accurate mass (mean m/z)
m (mean ppm)
Benfluorex Caffeine Clenbuterol Ephedrine/pseudoephedrine Fencamfamine Fenfluramine Mephentermine Phendimetrazine Phenmetrazine Phentermine Sibutramine Sildenafil Tadalafil Vardenafil
C19 H21 F3 NO2 C8 H11 N4 O2 C10 H19 Cl2 N2 O C10 H16 NO C15 H22 N C12 H17 F3 NO C11 H18 N C12 H18 NO C10 H16 NO C10 H16 N C17 H27 ClN C22 H31 N6 O4 S C22 H20 N3 O4 C23 H33 N6 O4 S
352.1519 195.0877 277.0869 166.1226 216.1747 232.1308 164.1434 192.1383 178.1226 150.1277 280.1827 475.2122 390.1448 489.2279
352.1515 195.0874 277.0866 166.1224 216.1744 232.1304 164.1432 192.1381 178.1224 150.1276 280.1822 475.2122 390.1449 489.2280
−1.1 −1.1 −1.0 −1.2 −1.2 −1.6 −1.2 −0.8 −1.2 −0.9 −1.6 0.1 0.1 0.3
Table 2 Limits of detection, limits of quantitation and linearity in the range from 50 to 2000 ng g−1 . Compound
LOD (ng g−1 )
LOQ (ng g−1 )
Slope ± SD
Benfluorex Caffeine Clenbuterol Fencamfamine Fenfluramine Mephentermine Phendimetrazine Phenmetrazine Phentermine Pseudoephedrine Sibutramine Sildenafil Tadalafil Vardenafil
10 1 1 1 1 10 5 5 25 1 10 25 5 25
50 50 50 50 50 50 50 50 50 50 50 50 50 50
0.0488 0.0250 0.2452 0.2210 0.0124 0.0662 0.0332 0.0637 0.0028 0.0192 0.0523 0.0660 0.0219 0.0013
relative intensities < 10% in the unknown and in the reference, and mass accuracies of characteristic ions should not vary more than 5 ppm. This method was applied to 36 nutritional supplements sold in fitness centers and in websites, collected from police officers in Italy. HRMS analysis showed the presence of ephedrine in three samples at a concentration ranging from 370 ng g−1 to 1000 ng g−1 (in one case ephedrine was not declared in the label), ephedrine and pseudoephedrine in one sample at 460 ng g−1 and 540 ng g−1 , respectively. A supplement, that was supposed to contain ephedrine according to the label, was negative. Moreover, caffeine, whose presence was reported in the label of six-supplement only, was detected at concentrations ranging from 225 up to 40500 ng g−1 in eight supplements. In the cases where caffeine
± ± ± ± ± ± ± ± ± ± ± ± ± ±
Intercept ± SD
0.0043 0.0009 0.0095 0.0151 0.0003 0.0053 0.0040 0.0055 0.0004 0.0026 0.0092 0.0110 0.0108 0.0004
0.1722 0.0367 −2.3048 −1.7875 −0.7341 −0.0631 2.0147 0.8948 0.4399 −0.0597 0.1722 −0.1139 −0.5413 −0.0155
± ± ± ± ± ± ± ± ± ± ± ± ± ±
r2 ± SD
0.6062 0.0041 9.0434 3.6341 0.7991 0.1601 1.2860 1.5011 0.3337 0.2609 0.4224 0.1306 0.5864 0.0159
0.9971 0.9905 0.9924 0.9926 0.9956 0.9955 0.9951 0.9943 0.9944 0.9959 0.9917 0.9969 0.9957 0.9987
± ± ± ± ± ± ± ± ± ± ± ± ± ±
0.0013 0.0009 0.0007 0.0016 0.0029 0.0010 0.0033 0.0005 0.0014 0.0030 0.0004 0.0009 0.0046 0.0004
concentration exceeded the upper level of the calibration curve, the quantification was obtained by analysing only one milligram of each supplement. In four batches of tablets called “Kamagra” sildenafil was identified and, after proper dilution, quantified at a mean concentration of 200 mg g−1 , corresponding to nearly 100 mg per tablet. The extracted ionic chromatograms and mass spectra of a nutritional supplement sample containing ephedrine at 460 ng g−1 , pseudoephedrine at 540 ng g−1 and caffeine at 17800 ng g−1 are shown in Fig. 2. In Fig. 3a it is represented the isotopic cluster corresponding to the raw formula of sildenafil (C22 H31 N6 O4 S), relative to a Kamagra tablet. Fig. 3b shows the comparison of fragments obtained by CID experiments from Kamagra tablets and a sildenafil standard.
Table 3 Accuracy (E%), precision (expressed as CV%), matrix effect (ME%) and recovery (R%) data for selected analytes. Compound
Benfluorex Caffeine Clenbuterol Fencamfamine Fenfluramine Mephentermine Phendimetrazine Phenmetrazine Phentermine Pseudoephedrine Sibutramine Sildenafil Tadalafil Vardenafil
50 ng g−1
500 ng g−1
2000 ng g−1
E%
CV%
ME%
R%
E%
CV%
ME%
R%
E%
CV%
ME%
R%
8 −3 11 9 3 −12 1 4 7 3 −9 3 3 1
5 8 12 12 7 7 8 10 12 3 5 1 2 2
73 86 83 96 95 97 80 75 90 74 91 93 99 98
24 42 36 44 33 33 33 27 36 29 32 34 36 44
6 13 7 3 −8 8 2 −2 1 −2 −7 −3 −1 −3
8 8 6 8 6 3 8 8 15 11 4 5 5 2
88 93 93 99 101 95 101 107 105 107 98 94 101 107
45 36 30 54 41 33 43 35 37 28 35 41 33 33
4 −2 −1 1 3 −2 3 1 −2 2 −3 1 3 2
3 5 2 3 2 1 3 4 5 4 1 3 1 0.4
88 93 90 96 95 92 100 95 97 96 103 91 108 94
36 39 49 42 45 34 35 29 42 29 42 41 44 39
148
S. Strano-Rossi et al. / Journal of Pharmaceutical and Biomedical Analysis 106 (2015) 144–152 12.22
tadalafil
100
Tadalafil
50
6.45
sibutramine
0 100 50 0.97
benfluorex
9.30
3.72
sildenafil vardenafil
6.18
fencamfa.. fenfluram.. Clenbuterol
NL: 2.92E5 m/z= 475.2074-475.2170 MS 100ng_g_smart_sto p_2_stimolanti
10.29
Sildenafil
50
2.95 3.31
4.41 4.94
7.40
5.84
11.24
NL: 3.73E4 m/z= 489.2230-489.2328 MS 100ng_g_smart_sto p_2_stimolanti
9.65
Vardenafil
50
9.80
NL: 1.82E6 m/z= 216.1725-216.1769 MS 100ng_g_smart_sto p_2_STIMOLANTI
8.81
Fencamfamine 50 8.95 8.15
9.21 10.26
11.12
NL: 6.61E5 m/z= 232.1285-232.1331 MS 100ng_g_smart_sto p_2_STIMOLANTI
8.80
Fenfluramine
50
9.28
0 100
7.39
Clenbuterol
50 0.63
0 100 Caffeine
11.33
8.62
0 100
5.13
6.42 6.66
8.28
9.04 9.36 9.88 10.34
11.58
12.30 12.85
Caffeine
50
0.14 0.40 1.26 1.95 2.29 2.68 3.26
3.90
5.43 5.98
4.67
6.76
10.12 10.21 10.28 11.32
7.38 7.74 8.40
13.54
NL: 4.59E5 m/z= 277.0841-277.0897 MS 100ng_g_smart_sto p_2_stimolanti
NL: 1.15E5 m/z= 195.0857-195.0897 MS 100ng_g_smart_sto 13.77 p_2_stimolanti
7.22
0 100 Phenterm..
NL: 4.56E5 m/z= 352.1484-352.1554 MS 100ng_g_smart_sto p_2_STIMOLANTI
Benfluorex
50
0 100
12.59
NL: 4.75E4 m/z= 150.1262-150.1292 MS 100ng_g_smart_sto 6.35 p_2_stimolanti 7.31 8.43 8.78 9.61 9.96 10.23 11.27 13.66 0.68 1.10 1.66 2.05 2.86 3.21 3.69 4.32 4.92 5.43 5.85 12.06 12.87 NL: 1.53E5 6.59 m/z= 164.1418-164.1450 MS 100ng_g_smart_sto p_2_stimolanti 6.91 7.49 7.88 8.60 9.13 1.39 3.09 12.02 12.83 13.64 9.89 10.61 11.08 NL: 1.25E7 6.31 m/z= 192.1364-192.1402 MS 100ng_g_smart_sto p_2_stimolanti 7.04 7.63 7.93 8.44 9.58 9.82 10.33 11.05 11.60 12.25 12.78 13.95 0.70 1.25 1.69 2.40 2.81 3.70 3.97 4.87 5.66 NL: 1.60E7 6.24 m/z= 178.1208-178.1244 MS 100ng_g_smart_sto p_2_stimolanti 7.01 7.63 7.97 8.52 9.14 9.59 10.00 10.54 11.13 12.16 12.86 0.71 1.14 2.01 2.50 3.41 3.83 4.37 4.75 5.53 13.95 NL: 1.28E5 5.24 0.68 m/z= 166.1209-166.1243 MS 100_ng_g_smart_st 5.42 op_2_stimolanti 1.07 5.66 6.22 7.14 1.25 2.31 7.91 8.22 0.23 3.29 4.05 4.62 4.84 9.30 9.79 10.09 10.96 11.79 12.79 13.42 7.04
Phentermine
50 0 100
Mephent..
12.09 12.68
10.06 11.12
0 100
Phendimetr..
NL: 7.52E5 m/z= 280.1799-280.1855 MS 100ng_g_smart_sto p_2_STIMOLANTI
Sibutramine
0 100
Phenmetraz..
13.59
11.61
0 100
Pseudoeph..
12.67
10.11
NL: 3.85E5 m/z= 390.1409-390.1487 MS 100ng_g_smart_sto p_2_stimolanti
Mephentermine
50 0 100
Phendimetrazine
50 0 100
Phenmetrazine
50 0 100
Pseudoephedrine
50 0 0
1
2
3
4
5
6
7 Time (min)
8
9
10
11
12
13
14
Fig. 1. HR ion chromatograms, based on accurate mass of each analyte, of a nutritional supplement sample spiked with the analytes of interest at a concentration of 100 ng g−1 .
S. Strano-Rossi et al. / Journal of Pharmaceutical and Biomedical Analysis 106 (2015) 144–152
149 NL: 2.58E8 m/z= 195.0857195.0897 MS 130624_ca mp3
7.12
100 90
Caffeine
80 Relative Abundance
70 60 50 40 30 20
7.47 7.92 8.37 8.74 8.99 9.48
10 0.06 0.49
0 0
1
2
3
4
5
6
7 Time (min)
8
9
10
11
12.67 13.17 13.52 13.92 12
MH+
80
experimental isotopic paern
60 40 20
13
14
NL: 1.87E8 130624_camp3#528-536 RT: 7.09-7.20 AV: 9 SB: 26 6.82-7.00 , 7.41-7.56 T: FTMS {1,1} + p ESI Full ms [80.00-500.00]
195.0875
100 Relative Abundance
10.32 10.72 11.19
5.65 5.99 6.43 6.71
196.0908 195.0132 195.1619 195.3110 195.0877
0 100
195.5315 195.6792
195.9329
197.0915 NL: 2.11E4
MH+
80
calculated isotopic paern
C 8 H 10 N 4 O 2 +H: C 8 H 11 N 4 O 2 p (gss, s /p:40) Chrg 1 R: 100000 Res .Pwr . @FWHM
60 40 20
196.0911 197.0919
0 195.0
195.2
195.4
195.6
195.8
196.0
196.2
196.4
196.6
196.8
197.0
197.2
m/z
90
5.06
Ephedrine Pseudoephedrine
80 Relative Abundance
NL: 1.10E5 m/z= 166.1209166.1243 MS 130624_ca mp3
5.22
100
70 60 50 40 30 20 10 0 .2 5
1.02
1.54
2.19
2.92 3.27
3 .9 2
5.46 5.63 6.13
4 .6 5
7.14 7.56 7.93
0 0
2
3
4
5
6
7 Time (min)
8
8.10
9.08 9.47 9.82 10.43 10.97 11.36 12.03 12.58 9
10
11
12
80
MH+ experimental isotopic paern
60 40 20
167.1256 166.8371 166.9987
0 100
13.32 13
14
NL: 4.64E4 130624_camp3#583-594 RT: 5.01-5.11 AV: 12 SB: 10 4.92-5.00 , 5.29-5.37 T: FTMS {1,1} + p ESI Full lock ms [50.00-550.00]
166.1224
100 Relative Abundance
1
167.5476
168.1277 NL: 2.09E4 C 10 H 15 NO +H: C 10 H 16 N 1 O 1 p (gss, s /p:40) Chrg 1 R: 100000 Res .Pwr . @FWHM
80
MH+ calculated isotopic paern
60 40 20
167.1260 168.1293
0 166.0
166.2
166.4
166.6
166.8
167.0
167.2 m/z
167.4
167.6
167.8
168.0
168.2
168.4
Fig. 2. HR ion chromatograms, based on accurate mass of each analyte, and experimental and calculated MH+ isotopic patterns of a nutritional supplement sample containing ephedrine at 460 ng g−1 , pseudoephedrine at 540 ng g−1 and caffeine at 17,800 ng g−1 .
150
S. Strano-Rossi et al. / Journal of Pharmaceutical and Biomedical Analysis 106 (2015) 144–152
Fig. 3. (a) [M + H]+ isotopic cluster of the component of a Kamagra tablet, corresponding to the raw formula of sildenafil (C22 H31 N6 O4 S); and (b) fragments obtained in CID experiments at 40 V for kamagra tablets (upper box) and sildenafil standard (lower box).
4. Discussion The identification of stimulants and PDE5Is in nutritional supplements is a matter of concern for the public health. In some cases they can be even not declared in the label, and this poses a risk to the
health, especially on subjects with pathologies that can be worsen by the above mentioned drugs, or on professional athletes, that can be exposed to an adverse analytical finding during antidoping tests. For instance, the presence of benfluorex, although not expressly included in the WADA list of prohibited substances, can lead to an
S. Strano-Rossi et al. / Journal of Pharmaceutical and Biomedical Analysis 106 (2015) 144–152
adverse analytical finding for norfenfluramine, its main metabolite. Other studied classes of substances, such as stimulants and anorectics, besides being included in the WADA prohibited list, may be dangerous for the health and for this reason some of them have been withdrawn from the market in many Countries [18–21]. Even when they are declared, such as in the case of kamagra containing sildenafil, these kinds of drugs, when prepared by unknown laboratories and sold on-line, are not subjected to the national and international controls, that are compulsory for pharmaceutical products, and could be potentially dangerous. A method for the identification of these drugs is therefore needed. The proposed approach provides the use of a sodium hydroxide purification step prior to the extraction using a mixture of solvents. With this preparation, a clean extract, even in the case of oily supplements or energy bars, can be obtained, and all the classes of analytes investigated are satisfactorily extracted. In one paper reporting the analysis of clenbuterol in tablets is described a similar approach with the use of diluted KOH for dissolving the sample and extraction with tertbutyl-methyl ether [3]. Other studies perform a simple dissolving of supplements in methanol [6,7,12] or acetonitrile/water [11] and a subsequent filtration, obtaining good recoveries for the analysis of anorectic compounds and PDE5Is; these approaches, however, cannot be applied in case of complex matrices such as bars or oily preparations. The optimized chromatographic gradient allows a good separation of analytes, also for similar compounds. For example, in the case of ephedrines, it was possible the discrimination of isomers (ephedrine/pseudoephedrine), as can be seen in Fig. 2. Moreover, the full scan acquisition enables the identification of other analytes not included in the design of method development. Indeed, it was possible to identify traces of norephedrine/norpseudoephedrine in the samples containing ephedrines. Both isomers were detected in the sample with ephedrine and pseudoephedrine, whilst only traces of norephedrine and methylephedrine were found in one of the samples showing higher amounts of ephedrine. On the other end, the mass spectrometric analysis performed at high mass resolution enhances the specificity of the method for the seeked compounds and enables the evaluation of the fine molecular structures, allowing the reliable identification of the elemental composition of analytes. In case of kamagra, for example, the use of a resolution of 100,000 (FWHM at m/z 200) allows to evince the presence of a sulfur atom, as it is possible to observe a splitting of the [M+2] isotopes (see Fig. 3a). The resolved signal at m/z 477.2080 in fact corresponds to the contribute of the isotope 34 S, which gives a theoretical relative mass with +1.996 with respect to the M+0 monoisotopic ion (m/z 475.2122), while the second peak at m/z 477.2181 corresponds to the contribute of 13 C2 .
5. Conclusions This study reports the application of a method for the analysis of dietary supplements for critical compounds potentially dangerous for the public health, especially when they are not reported in the labels. By using the benefits of the versatility of liquid chromatography coupled to the high resolution mass spectrometry, a screening method was developed and validated for the analysis of stimulants, anorectics and PDE5Is in nutritional supplements. Cheap and efficient sample preparation step enabled an effective purification of the supplements, and HRMS enables high selectivity for the monitoring of compounds that share the same nominal mass but differ in their elemental composition and therefore have a different exact mass, and therefore allows to distinguish between analytes of interest and isobaric contaminants coming from a
151
complex matrix such as nutritional supplements. Moreover, high resolution mass spectrometry allows the accurate identification of analytes, through the exact masses of target compounds and the associated isotopic patterns. The great potentiality of high resolution mass spectrometry technique is also related to the possibility of reprocessing previously acquired data files when a further substance has to be investigated. The application of the method to the analysis of 36 dietary supplements, sold in fitness centers and in unauthorized online markets, allowed the detection of substances not declared in the products labels, such as ephedrine and caffeine. This highlights the importance of the analysis of dietary supplements as they are highly consumed and sometimes misused by general public unaware of the potential health risk. Acknowledgements This work was carried out in the framework of the project “Smart Stop” – Control on the sites selling products dangerous for the public health, financed by the Antidrug Policies Department, Presidency of Italian Council of Ministers. References [1] Directive 2002/46/EC, http://ec.europa.eu/food/food/labellingnutrition/ supplements/index en.htm (accessed 20.04.14). [2] Directive 2010/84/EU, http://ec.europa.eu/health/files/eudralex/vol-1/dir 2010 84/dir 2010 84 en.pdf (accessed 20.04.14). [3] M.K. Parr, K. Koehler, H. Geyer, S. Guddat, W. Schanzer, Clenbuterol marketed as dietary supplement, Biomed. Chromatogr. 22 (2008) 298–300. [4] J. Jung, M. Hermanns-Clausen, W. Weinmann, Anorectic sibutramine detected in a Chinese herbal drug for weight loss, Forensic Sci. Int. 161 (2006) 221–222. [5] C. Vidal, S. Quandte, Identification of a sibutramine-metabolite in patient urine after intake of a “pure herbal” Chinese slimming product, Ther. Drug Monit. 28 (2006) 690–692. [6] P. Zou, S.S. Oh, K.H. Kiang, M.Y. Low, B.C. Bloodworth, Detection of sibutramine, its two metabolites and one analogue in a herbal product for weight loss by liquid chromatography triple quadrupole mass spectrometry and time-of-flight mass spectrometry, Rapid Commun. Mass Spectrom. 21 (2007) 614–618. [7] H.J. Kim, J.H. Lee, H.J. Park, S.H. Cho, S. Cho, W.S. Kim, Monitoring of 29 weight loss compounds in foods and dietary supplements by LC–MS/MS, Food Addit. Contam. A: Chem. Anal. Control Expo. Risk Assess. 31 (2014) 777–783. [8] F.T. Delbeke, Nutritional supplements and doping, in: C. Peters, T. Schulz, H. Michna (Eds.), Biomedical Side Effects of Doping, Cologne, 2001, pp. 155–161. [9] O. de Hon, B. Coumans, The continuing story of nutritional supplements and doping infractions, Br. J. Sports Med. 41 (2007) 800–805, discussion 805. [10] C.N. Man, N.M. Nor, R. Lajis, G.L. Harn, Identification of sildenafil, tadalafil and vardenafil by gas chromatography–mass spectrometry on short capillary column, J. Chromatogr. A 1216 (2009) 8426–8430. [11] F. Shi, C. Guo, L. Gong, J. Li, P. Dong, J. Zhang, P. Cui, S. Jiang, Y. Zhao, S. Zeng, Application of a high resolution benchtop quadrupole–Orbitrap mass spectrometry for the rapid screening, confirmation and quantification of illegal adulterated phosphodiesterase-5 inhibitors in herbal medicines and dietary supplements, J. Chromatogr. A. 1344 (2014) 91–98. [12] H.M. Lee, B.J. Lee, A novel approach to simultaneous screening and confirmation of regulated pharmaceutical compounds in dietary supplements by LC/MS/MS with an information-dependent acquisition method, Food Addit. Contam. A: Chem. Anal. Control Expo. Risk Assess. 28 (2011) 396–407. [13] C.M. Gryniewicz, J.C. Reepmeyer, J.F. Kauffman, L.F. Buhse, Detection of undeclared erectile dysfunction drugs and analogues in dietary supplements by ion mobility spectrometry, J. Pharm. Biomed. Anal. 49 (2009) 601–606. [14] B.J. Venhuis, D. de Kaste, Towards a decade of detecting new analogues of sildenafil, tadalafil and vardenafil in food supplements: a history, analytical aspects and health risks, J. Pharm. Biomed. Anal. 69 (2012) 196–208. [15] Wada, List of Prohibited Substances and Methods, 2014, http://www.wadaama.org/Documents/World Anti-Doping Program/WADP-Prohibited-list/ 2014/WADA-prohibited-list-2014-EN.pdf (accessed 20.04.14). [16] S. Martello, M. Felli, M. Chiarotti, Survey of nutritional supplements for selected illegal anabolic steroids and ephedrine using LC-MS/MS and GC–MS methods, respectively, Food Addit. Contam. 24 (2007) 258–265. [17] B.K. Matuszewski, M.L. Constanzer, C.M. Chavez-Eng, Strategies for the assessment of matrix effect in quantitative bioanalytical methods based on HPLC–MS/MS, Anal. Chem. 75 (2003) 3019–3030.
152
S. Strano-Rossi et al. / Journal of Pharmaceutical and Biomedical Analysis 106 (2015) 144–152
[18] A.J. Scheen, Cardiovascular risk-benefit profile of sibutramine, Am. J. Cardiovasc. Drugs 10 (2010) 321–334. [19] C. Tribouilloy, S. Marechaux, Y. Jobic, A. Jeu, S. Ederhy, E. Donal, P. Reant, E. Arnalsteen, J. Boulanger, T. Garban, P.V. Ennezat, M. Andrejak, D. Rusinaru, Frequency of drug-induced valvular heart disease in patients previously exposed to benfluorex: a multicentre prospective study, Eur. Heart J. 34 (2013) 3580–3587.
[20] M. Andrejak, C. Tribouilloy, Drug-induced valvular heart disease: an update, Arch. Cardiovasc. Dis. 106 (2013) 333–339. [21] L. Abenhaim, Y. Moride, F. Brenot, S. Rich, J. Benichou, X. Kurz, T. Higenbottam, C. Oakley, E. Wouters, M. Aubier, G. Simonneau, B. Begaud, Appetite-suppressant drugs and the risk of primary pulmonary hypertension, N. Engl. J. Med. 335 (1996) 609–616.
