Doi:10.1016/j.chroma.2005.01.038

Journal of Chromatography A, 1066 (2005) 89–95 Simultaneous determination of sildenafil, vardenafil and tadalafil as forbidden components in natural dietary supplements for male sexual potency by high-performance liquid chromatography–electrospray Xiaolan Zhu, Song Xiao, Bo Chen, Fei Zhang, Shouzhuo Yao, a Key Laboratory of Chemical Biology and Traditional Chinese Medicine Research, Ministry of Education, Hunan Normal University, Changsha 410081, China b National Institute for Nutrition and Food Safety, Chinese Center for Disease Control and Prevention, Beijing 100021, China Received 17 February 2004; received in revised form 17 December 2004; accepted 7 January 2005 Abstract
A high-performance liquid chromatographic method coupled with ultraviolet detection and electrospray ionization mass spectrometry (HPLC–UV–ESI-MS) was developed for simultaneous determination of banned additives—sildenafil, vardenafil and tadalafil in dietarysupplements for male sexual potency. The separation was achieved on a C18 column with acetonitrile and aqueous solution (20 mmol ammoniumacetate, 0.2% formic acid) as mobile phase at a flow rate of 1 ml/min with a linear gradient program. UV detection was at 292 nm. Identificationof drugs was accomplished using ESI-MS. Good linearity between response (peak area) and concentration was found over a concentrationrange of 0.8–80 ␮g/ml for sildenafil; 2.25–225 ␮g/ml for vardenafil; and 1.1–110 ␮g/ml for tadalafil, with regression coefficient is better than0.999. The recovery of the method ranged from 93.3 to 106.1%, and the relative standard deviation varied from 2.0 to 5.6% (n = 6). Themethod has been successfully applied to the analysis of practical samples of natural dietary supplements.
2005 Elsevier B.V. All rights reserved.
Keywords: Sildenafil; Vardenafil; Tadalafil; Dietary supplements 1. Introduction
instruction because their over-dose might cause a series ofside-effects. For example, there were reports that color dis- Sildenafil (Viagra), an inhibitor of phosphodiesterase type crimination error scores increased after taking sildenafil 5 (PDE5), which was used in the past to treat patients with Tadalafil and vardenafil are safer than sildenafil, but they still pulmonary artery hypertension was approved for the can cause headache, dyspepsia and back pain treatment of erectile dysfunction (ED) in man by the US A dietary supplement is a product taken by mouth that Food and Drug Administration (FDA). Afterwards, varde- contains a “dietary ingredient” intended to supplement the nafil and tadalafil was also approved for the treatment of ED diet. The “dietary ingredients” in these products may include: These drugs should be administrated under doctors’ vitamins, minerals, herbs or other botanicals, amino acids,and substances such as enzymes, organ tissues, glandulars,and metabolites. The dietary supplement manufacturer is re- ∗ Corresponding author. Tel.: +86 731 8865515; fax: +86 731 8865515.
sponsible for ensuring that a dietary supplement is safe be- Co-Corresponding author.
E-mail addresses: dr-chenpo@vip.sina.com (B. Chen), fore it is marketed. The FDA is responsible for taking action against any unsafe dietary supplement product after it reaches 0021-9673/$ – see front matter 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.chroma.2005.01.038 X. Zhu et al. / J. Chromatogr. A 1066 (2005) 89–95 the market. In general, natural dietary supplements for male tadalafil is seldom reported, simultaneous determination of sexual potency consist of different herbal extracts such as these three analytes has been seldom reported up-to-date. The ginseng root (Panax ginseng C.A. Mey), lychee seed (Litchi purpose of this study was to develop a method for determining chinensis Sonn.), barbary wolfberry fruit (Lycium barbarum sildenafil, vardenafil and tadalafil simultaneously in natural L.), longan aril (Dimocarpus Longan Lour.), aweto (Cordy- dietary supplements. The structures of these compounds are ceps sinensis (Berk.) Sacc.), common peony root (Paeonia shown in The developed method showed some mer- lactiflora Pall.), Chinese magnoliavine fruit (Schisandra chi- its such as specificity, sensitivity, and simplicity in sample nensis (Turcz.) Baill), Indian bread (Poria cocos (Schw.) Wolf), shorthorned epimedium root (Epimedium brevicor-num Maxim.) and so on. These dietary supplements couldimprove male sexual potency without causing any danger, 2. Experimental
even when over dose occurs. However, in the SoutheasternAsian market, for the sake of profit, illegal dealers add some drugs such as sildenafil, vardenafil and so on to their prod-ucts. The illegal products may endanger people’s health. To The HPLC system used was a Waters (Milford, MA, USA) ensure the quality of this kind of dietary supplements and Alliance 2695 module, which was interfaced to a Waters 2487 protect people’s health, it is important to develop a method dual absorbance detector. The mass spectrometer used was a Micromass ZQ 2000 (Manchester, UK) equipped with an ESI Concerning the analysis of these compounds, there are a probe and quadrupole analyzer. The control of system and few reports that introduced the strategy for the determination data acquiring was performanced with Masslynx3.5 worksta- of sildenafil by the widely used HPLC technology Tracqui and Ludes developed an HPLC–MS method for the The standards of sildenafil and tadalafil were obtained determination of sildenafil Li et al. reported a method from Hunan Chemicals and Reagent Corp. (Changsha, for determining sildenafil with capillary electrophoresis China). Vardenafil (>98%, HPLC) was prepared in this labo- While the strategy for the determination of vardenafil and ratory on Waters preparative liquid chromatography of Prep Fig. 1. The structure of the investigated drugs.
X. Zhu et al. / J. Chromatogr. A 1066 (2005) 89–95 Table 1Main herbal constituents contained in samples Barbary wolfberry fruit, ginseng root, Chinese magnoliavine fruit Barbary wolfberry fruit, common peony root, Indian bread Barbary wolfberry fruit, shorthorned epimedium root Chinese magnoliavine fruit, barbary wolfberry fruit, ginseng root LC 4000 module. Samples for examination were purchased from supermarket (Changsha, China). All of these productsexamined are natural dietary supplements for male sexual The separation of the drugs was completed on a spherigel health, not for therapy of ED. The drugs are forbidden to be analytical column (Johnson, Dalian, China), which was added in these products according to the Chinese law. And packed with 5 ␮m C18 sillica. The mobile phase consisted these products are also not sexual potency enhancing prepa- of acetonitrile (A) and aqueous solution (B) containing ration. HPLC-grade acetonitrile and methanol were from 20 mmol/l ammonium acetate and 0.2% formic acid (v/v).
Shanghai Ludu Chemical Plant (Shanghai, China). Ultrapure The gradient elution was programmed as follows: A was water was prepared using a Millipore Milli-Q purification maintained at 35% within the first 10 min, then linearly in- system (Millipore, Bedford, MA, USA). Other reagents were creased to 80% during the following 5 min, then A maintained of analytical grade, including ammonium acetate and formic at 80% for another 5 min. The column was washed with 100% acid, triethylamine. Mobiles used for HPLC were filtered acetonitrile for 5 min after gradient elution, and then equili- (0.45 ␮m) and ultrasonically degassed before use.
brated for 10 min with the initial mobile phase for the nextinjection. The flow rate was kept at 1 ml/min and the columntemperature was maintained at 30 ◦C. Injection volume was 5 ␮l. The detection wavelength was set at 292 nm. The out-let of the UV detector was split, and only 0.2 ml/min portion Stock solutions of sildenafil, vardenafil and tadalafil were of the column effluent was delivered into the ion source of prepared in methanol. Their concentrations were 0.80, 2.25 and 1.10 mg/ml, respectively. One milliliter aliquots of each Electrospray was operated in positive ion mode to generate stock solution were transferred into a 10-ml volumetric flask, protonated ions and sodiated ions. The voltage of capillary, mixed and diluted to volume to yield a mixed standard so- extractor and RF lens was set at 3.2 kV, 4 and 0.5 V, respec- lution. Then, 5, 2, 1, 0.5, and 0.1 ml of the mixed standard tively. The temperature was maintained at 105 and 200 ◦C solution were transferred to five 10-ml volumetric flasks, and for source and desolvation, respectively. The gas flow rate diluted to volume with methanol to yield a series of work- for desolvation and cone was set at 250 and 50 l/h, respec- ing solutions. All stocking solutions and working solutions tively. The full scan mass spectra was acquired over a range of were stored in a refrigerator and brought to room temperature m/z 160–600. The cone voltage was switched from 60 to 20 V in scan mode at the point of 10 min according to the electricalstability of the drugs. In selective ionization recording (SIR), the cone voltages for sildenafil, vardenafil and tadalafil wereset at 50, 50, and 20 V, respectively.
Because the drugs have good solubility in water or methanol, they are often been added into fluid products suchas wine, beverage, and oral liquid formulation, etc. Hence, 2.5. Linearity, limit of detection, limit of quantification eight liquid products for examination (five oral liquid formu-lation, two wines, and one beverage, their herbal constituents The mixed standard solutions (the working solutions) at are listed in was purchased from a supermarket. The each concentration level were injected in triplicate, calibra- oral liquid formulation sample was filtered through a 0.45 ␮m tion curves were constructed by plotting the average peak nylon membrane, and 1 ml of the sample transferred into 50- areas of the standard compounds against the corresponding ml volumetric flask and diluted to volume with methanol.
concentrations. The limit of detection (LOD) of UV detection Then aliquot of the diluted solution was injected into the and MS–SIR was evaluated as the mass giving a signal equal HPLC–MS system. The wine sample and beverage sample to three times of noise (S/N = 3), the limit of quantification were just filtered off and injected into the HPLC–MS system (LOQ) was determined as the mass giving a signal equal to X. Zhu et al. / J. Chromatogr. A 1066 (2005) 89–95 3. Results and discussion
while sildenafil and vardenafil have a relatively stable struc-ture, so they can bear higher voltage. They gave little frag- ment ions under 50 V cone voltage, and produced only a fewfragment ions under 60 V. Tadalafil is easier to be cracked Firstly, methanol was applied to separate the tested com- down, the abundance of its molecular ion was still low even pounds, however, sildenafil and vardenafil could not be sepa- when the applied cone voltage is higher than 30 V. There- rated under the use of a mixed methanol aqueous solution with fore, as described in the previous experimental section, in any proportion of organic to aqueous phase. When acetoni- SIR mode, the cone voltage for sildenafil and vardenafil was trile was used, the two substances could be separated, their set at 50 V, while the cone voltage for tadalafil was set at 20 V; retention time and separation resolution mainly depended on in scan mode, the cone voltage was set at 60 V in the previ- the concentration of acetonitrile in the aqueous solution. A ous 10 min to generate some fragment ions for identification mobile phase consisting of acetonitrile–water programmed as of sildenafil and vardenafil, then switched to 20 V during the described in experimental section provided the best compro- mise between the separation efficiency and the time durationof the analytical procedure.
3.3. HPLC–UV–MS analysis of standards The examined compounds in this work all contain sev- eral N atoms in their structure; it results in serious peak- The examined analytes was baseline separated under the tailing on RP-C18 column if no modifier was added to the given chromatographic condition. shows the chro- mobile phase. In order to suppress peak-tailing, the effects of matogram of mixed standards recorded with 292 nm and with several additives and their concentration were investigated.
SIR, the retention times for sildenafil, vardenafil, tadalafil In liquid chromatography, triethylamine was the most com- are 7.9, 8.8, and 14.8 min, respectively. the mon additive used in analyzing compounds containing N mass spectrum of the three compounds. exhibites the atoms. In this work, 5, 10, 15, 25, and 50 mmol/l concen- intensive protonated molecule of sildenafil [M + H]+ at m/z trations of triethylamine were tested. It was found that when 475, m/z 497 is the sodiated molecule [M + Na]+ of sildenafil, 10 mmol/l triethylamine was employed, the peak is sharp and m/z 311 and 283 are the fragment ions of sildenafil. The as- relatively symmetric. When higher concentrations of triethy- signment can be done as follows: m/z 311 is the fragment lamine were used, the resulted peak shape was not improved ion losing an −R1 group, m/z 283 is the fragment ion los- any more, however the baseline shifted greatly when gradient ing an [R1 + ethyl] group. The same results were obtained mobile phase was employed and the resolution of sildenafil by Weinmann et al. Walker et al. The pres- and vardenafil decreased. And the ionization of all analytes ence of m/z 489 in represents the molecular ion was greatly suppressed, sildenafil and tadalafil gave no signal [M + H]+ of vardenafil, m/z 311 and 283 are the fragment even in the SIR chromatogram, and the signal of vardenafil ions losing an −R2 group and [R2 + ethyl] group, respec- was very weak. In addition, the effect of ammonium acetate, tively. And the ion at m/z 390 in is the molecular as a modifier of the mobile phase, was also investigated. When ion [M + H]+ of tadalafil, m/z 412 is the sodiated molecule 20 mmol/l ammonium acetate was used, the peak area RSD [M + Na]+, while m/z 268 is the result of losing an R3 group.
of three consecutive injections for each compound was less It can be seen from and B that sildenafil and varde- than 5% which is lower than in the case of triethylamine used nafil produce the same fragment ions. This is because that as modifier. However, when 50 mmol/l ammonium acetate they possess very similar structures and it can partly explain was used, the response of sildenafil; vardenafil and tadalafil why the two substances cannot be separated with methanol decreased 10.5, 18.4, and 20.4%, respectively, compared to that when 20 mmol/l ammonium acetate was applied. So, highconcentration of modifier was not recommended.
3.4. Linearity, limit of detection, limit of quantification Linearity of the three analytes was obtained over concen- tration range from 0.8 to 80 ppm, 2.25 to 225 ppm and 1.1 to The MS parameters were optimized attentively by flow 110 ppm, for sildenafil, vardenafil and tadalafil, respectively.
injection analysis (FIA). ESI is a soft ionization technique, Results are shown in these substances have con- Table 2Linearity, limit of detection (LOD), limit of quantification (LOQ) (n = 3) a Result with detection at 292 nm.
b Result with SIR.
X. Zhu et al. / J. Chromatogr. A 1066 (2005) 89–95 Fig. 2. The chromatogram of mixed standards. Peak identification: sildenafil(tR = 7.9), vardenafil (tR = 8.8) and tadalafil (tR = 14.8). The concentration ofthe three compounds in the mixture was 16, 45, and 22 ␮g/ml, respectively.
jugated structures and displayed intensive ultraviolet absorp- Fig. 3. The mass spectrum of examined analytes. (A) Sildenafil, (B) varde- tion, which resulted in quite a low LOD and LOQ with UV detection. Hence the UV detection method can be used forconventional analysis of these compounds even without mass spectrometry. However, the MS LOD of these compoundswas found to be even much more lower. On line analysis Precision of the method was evaluated by six consecu- displays that the proposed HPLC–ESI-MS method is advan- tive injections of the investigated samples, the resulting RSD tageous in trace analysis of these compounds and can provide structure information for identification when no standards are The accuracy of the method was studied by calculating the mean recovery of the target compounds after adding stan- X. Zhu et al. / J. Chromatogr. A 1066 (2005) 89–95 Table 3Precision and recoveries (n = 3) The result was obtained by employing UV detection at 292 nm.
dards to three blank samples (wine, beverage, oral liquid 94.0–106.1%. These results about precision and accuracy met formulation) at low, medium and high levels. Each sample of the same concentration was injected at least three times.
The results are summarized in From this Table, it 3.6. HPLC–UV–MS analysis of samples can be seen, that the mean recovery for all three drugs was Herbs are very complex because they contain many kinds of compounds. The samples examined in present work in-cluded barbary wolfberry fruit, ginseng root, Chinese mag-noliavine fruit, aweto, Indian bread, common peony root,shorthorned epimedium root and lychee seed. The com-positions of all these herbs are rather complicate. How- Fig. 4. The chromatogram of sample (oral liquid formulation 5) acquiredwith detection at 292 nm. (A) Chromatogram after adding three standards,(B) chromatogram of sample before adding standards. Peak identification: Fig. 5. The chromatogram of oral liquid preparation sample 2. (A) Recorded 1, sildenafil; 2, vardenafil and 3, tadalafil.
with detection at 292 nm; (B) recorded with SIR 489.
X. Zhu et al. / J. Chromatogr. A 1066 (2005) 89–95 ever, under the above-given conditions, no interference References
from these herbs was observed. shows the chro-matogram of oral liquid formulation 5 after (A) and before [1] N. Danchin, Ann. Cardiol. Angeiol. (Paris) 51 (2002) 341.
(B) adding sildenafil, vardenafil and tadalafil standards. It [2] B.K. Sastry, C. Narasimhan, N.K. Reddy, B. Anand, G.S. Prakash, P.R. Raju, D.N. Kumar, Indian Heart J. 54 (2002) 410.
can be seen that no interfering components were co-eluted [3] S.S. Kothari, B. Duggal, Indian Heart J. 54 (2002) 404.
with these three drugs simultaneously. Among the eight ex- [4] H. Porst, Int. J. Impot. Res. Suppl. 1 (2002) 57.
amined samples, one sample (oral liquid fomulation 2) was [5] J. Kuan, G. Brock, Expert Opin. Invest. Drugs 11 (2002) 1605.
found to contain vardenafil, its concentration was 2.25 mg/ml [6] T.J. McCulley, J.K. Luu, M.F. Marmor, W.J. Feuer, Ophthalmologica (RSD = 1.7%, n = 6). The chromatogram of this sample is [7] W.J. Hellstrom, J.W. Overstreet, A. Yu, K. Saikali, W. Shen, C.M.
Beasley, V.S. Watkins, J. Urol. 170 (2003) 887.
[8] L.A. Hicklin, C. Ryan, D.K. Wong, A.E. Hinton, J.R. Soc. Med. 95 4. Conclusion
[9] V. Nagaraju, D. Sreenath, J.T. Rao, R.N. Rao, Anal. Sci. 19 (2003) With the improvement in production technology of silde- [10] J.Y. Cho, H.S. Lim, K.S. Yu, H.J. Shim, I.J. Jang, S.G. Shin, J.
nafil and its analogous compounds, the quantity of these com- pounds is becoming bigger and bigger and their prices are [11] J.D. Cooper, D.C. Muirhead, J.E. Taylor, P.R. Baker, J. Chromatogr.
on decline, hence even more of these compounds are being [12] J. Lia, T.W. Chang, J. Chromatogr. B 765 (2001) 161.
added to dietary supplements by illegal businessmen. The [13] M.T. Sheu, A.B. Wu, G.C. Yeh, A. Hsia, H.O. Ho, J. Lia, T.W.
method presented in this paper is useful for simultaneous Chang, J. Chromatogr. B 791 (2003) 255.
determination of sildenafil, vardenafil, tadalafil. It can be [14] V. Nagaraju, D. Sreenath, J.T. Rao, R.N. Rao, Anal. Sci. 19 (2003) employed to inspect those dietary supplements which may [15] N.D. Dinesh, B.K. Vishukumar, P. Nagaraja, N.M. Made Gowda, contain these substances to ensure people’s safety, and the K.S. Rangappa, J. Pharm. Biomed. Anal. 29 (2002) 743.
suggested method has the advantage of simplicity, rapidity [16] E. Angela, A. Tom, N.D. Weng, J. Chromatogr. B 768 (2002) [17] A. Tracqui, B. Ludes, J. Anal. Toxicol. 27 (2003) 88.
[18] R.K. Li, T. Bo, H.W. Liu, K.A. Li, Se Pu 20 (2002) 335.
Acknowledgement
[19] W. Weinmann, M. Bohnert, A. Wiedemann, M. Renz, N. Lehmann, S. Pollak, Int. J. Legal Med. 114 (2001) 252.
This work was financially supported by the Na- [20] D.K. Walker, M.J. Ackland, G.C. James, G.J. Muirhead, D.J. Rance, P. Wastall, P.A. Wright, Xenobiotica 29 (1999) 297.
tional Foundation of Key Technologies for Food Safety,China

Source: http://www.ezkaoyan.com/image/html/20071210145002.pdf