Journal of Analytical Toxicology 2015;39:130 –132 doi:10.1093/jat/bku140 Advance Access publication December 11, 2014
Article
Confirmation and Quantification of Clenbuterol in Horse Urine Using Liquid Chromatography Tandem Mass Spectrometry Triple Quadrupole Jennifer Bishop*, Brendan Heffron, Lisa Taddei, Marc Benoit, Laura Hurt, Sara Costello, Melissa Gross and Adam Negrusz Animal Forensic Toxicology Laboratory, Department of Biopharmaceutical Sciences, College of Pharmacy, University of Illinois at Chicago, 2242 West Harrison Street, Chicago, IL 60612, USA *Author to whom correspondence should be addressed. Email: [email protected]
Clenbuterol (CLE) is used in horses as a bronchodilator and for its anabolic steroid-like effects. CLE is a Class 3 drug according to current Association of Racing Commissioners International (ARCI) Uniform Classification Guidelines. The Racing Medication and Testing Consortium recommended a urine CLE threshold of 140 pg/mL after careful scientific review of the results of studies describing the disposition of CLE in the horse and this threshold was adopted by the ARCI. Enzyme-linked immunosorbent assay was previously used to screen samples for CLE in Illinois, but could not detect such low concentrations in urine. Thus, a liquid –liquid extraction of CLE from urine followed by quantification by liquid chromatography – tandem mass spectrometry was developed and validated. Method validation included testing stability, ion suppression and enhancement, precision, accuracy and uncertainty. Intra-, interday and total precision and accuracy were calculated for each control and found to be within the + 15% acceptance range. The Guide to the Expression of Uncertainty in Measurement approach was used to calculate uncertainty, which was 11% at the 95% confidence level. In the past 5 years, only 15 samples were reported as positive for CLE in Illinois. This new method was used in a pilot program to screen and confirm samples received from thoroughbred and harness horses.
Introduction Clenbuterol (CLE) is a bronchodilator and is the only b2 agonist approved by the Federal Drug Administration for use in horses. However, CLE can alter performance in race horses, including control of bronchospasms and anabolic effects. Horses that have been given chronic administration of CLE have shown significant decreases in percent fat and fat mass, and increases in fatfree mass (1 –3). VentipulminTM syrup containing CLE hydrochloride is approved by the US Food and Drug Administration for use in horses to treat inflammatory airway disease and chronic obstructive pulmonary disease. Anecdotal reports suggest that some trainers may administer CLE in various formulations in an attempt to achieve an anabolic effect particularly since the introduction of rules of racing that prohibit the use of anabolic steroids. CLE is a Class 3 drug according to current Association of Racing Commissioners International Uniform Classification Guidelines (4). Due to the fact that CLE is used therapeutically, the Racing Medication and Testing Consortium (RMTC) suggested a threshold, which has been accepted by the Illinois Racing Board (5). Previous research showed that the lowest effective dose of CLE is 0.8 mg/kg. Multiple methods of administration have been tested for efficacy, and the most effective ones were oral and intravenous (i.v.) administration (1). CLE given by intratracheal injection
showed no effect on horses that had decreased lung function, and therefore, is not considered an effective administration of the drug (6). Although oral administration was shown to have an effect for 1 h longer than i.v. administration, there is no pharmacokinetic difference between the two routes of administrations (1, 7). After 22 horses were given an oral dose of 0.8 mg/kg bwt twice a day for a total of 30 days, the highest concentration of CLE in urine found 14 days postadministration was 140 pg/mL (8) and it was chosen by the RMTC as the recommended threshold. Enzyme-linked immunosorbent assay (ELISA) was previously used as the screening method for CLE in Illinois based on local decisions. Even though the manufacturer’s limit of detection (LOD) is 1 ng/mL and previous reports have shown the LOD to be 0.1 ng/mL, the LOD of the University of Illinois at Chicago Animal Forensic Toxicology Laboratory was established to be 5 ng/mL (9). Thus, a method based on liquid –liquid extraction of CLE and its internal standard from urine followed by quantification by liquid chromatography mass spectrometry (LC – MS) was developed and validated (8). The primary objective of this study was to develop a method to confirm and quantify CLE in race horses. Due to the much lower screening levels, a pilot program was put into place for 3 months in order to determine how many horses would test positive over the new threshold and as a result, the violations would have been reported. Experimental Pilot program During the time before the threshold was implemented, a pilot program was performed in order to determine how many samples would have been reported over the threshold. A total of 577 samples (304 harness and 273 thoroughbred horses) were screened for CLE. This was only a small percentage of the total number of samples coming into the laboratory. CLE administration and sample collection The administration was provided by Interstate Drug Testing Alliance and Iowa State University. Two mare, quarter horses were used in this study. CLE hydrochloride was prepared by dissolving 2.59 mg in 5 mL of sterile saline. The solution was then filtered. The lowest therapeutic dose of 0.8 mg/mL was injected intravenously based on their weights of 643 and 606 kg. Predose urine was collected from both horses and samples were then collected at 2, 4, 8, 24 and 48 h postdose. Urine was collected via Foley catheter from both horses, and was pooled at each time collection.
# The Author 2014. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]
Reference compounds CLE hydrochloride and clenbuterol-d 9 (CLE-d 9) hydrochloride powders were obtained from Sigma-Aldrich (St. Louis, MO, USA). Stock solutions were made at 1 mg/mL in methanol for both CLE and the internal standard. Working solutions were made by diluting the stock solutions to a concentration of 1 mg/mL.
The parent ions that were used for detecting CLE and CLE-d 9 (CLE-d 9) were m/z 277.1 and 286.2, respectively. Product ions for CLE were 203, 168 and 132.1. The most abundant product ion was 203 and was therefore used for quantifying. CLE-d 9 had product ions of 204 and 66.2.
Method validation Reagents Methyl tert-butyl ether, methanol, formic acid, acetonitrile, water, glacial acetic acid and potassium carbonate were obtained from Fisher Scientific (Hanover Park, IL, USA), and were HPLC grade or better. Sodium acetate trihydrate was obtained from Mallinckrodt Pharmaceuticals (St. Louis, MO, USA). b-Glucuronidase was obtained from Campbell Science (Rockford, IL, USA). Blank equine urine was obtained from Bioreclamation LLC (Westbury, NY, USA).
Standard curve and quality control preparations Six curve points were made in blank equine urine at concentrations of 100, 200, 400, 800, 1600 and 3200 pg/mL. Quality control preparations were made at two concentrations in blank equine urine (2000 and 250 pg/mL).
Extraction method For all blank, quality control, curve points and samples, 2 mL of equine urine was placed in screw cap tubes. To each sample, 500 mL of 0.9 M sodium acetate buffer containing internal standard at a concentration of 500 pg/mL and 50 mL of bglucuronidase were added. The samples were mixed, capped loosely and incubated overnight at 378C. After cooling, 2 mL of 0.5 M potassium carbonate and 5 mL of methyl tert-butyl ether were added to each sample. The tubes were capped, mixed by rotorack for 10 min and centrifuged for 10 min at 3000 rpm. The solvent layer was then transferred to a clean tube and evaporated to dryness under nitrogen in a water bath at 378C. Two hundred and fifty microliters of 0.2% formic acid in water was added to each tube. The samples were vortexed and transferred to an HPLC autosampler deep well. Twenty microliters from each sample were then injected onto the LC –MS-MS.
Instrumentation An Agilent 1200 Series HPLC coupled with an Agilent 6410 Series Triple Quadrupole Mass Spectrometer (LC – MS-QQQ) in positive ionization mode was used to detect and quantify CLE. The column used was an Agilent Poroshell 120 EC-C 18 2.1 i.d.100 mm, 2.7 mm particle diameter. The pump flow was set to 500 mL/min and the column temperature was kept at 308C. Solvent A was 0.2% formic acid in water and Solvent B was 90 : 10 acetonitrile : water. The initial concentration of Solvent B was 10% and ramped to 45% over 2 min. Then the concentration was increased to 100% over 1 min and held for 2 min before re-equilibrating to 10%. Gas temperature was set to 3508C while the gas flow was at 11 L/min. Ultra-high-purity nitrogen was used for the drying, sheath and collision gases. The nebulizer pressure was at 60 psi and the capillary voltage was 3500 V.
Method validation included selectivity, carryover, LOD, LLOQ, stability, extraction recovery, matrix effect, process efficiency, accuracy and precision. Selectivity was determined by analyzing three blank equine urine samples from three different sources. These samples were extracted and injected onto the LC – MS-MS to analyze for the presence of CLE and CLE-d 9 product ions based on injected standard retention times. Carryover and contamination were assessed by analyzing three blank samples that were injected after the highest concentration curve point. LOD was determined by finding the lowest concentration where the signal to noise (S/N) of any product ion was 5 : 1. LLOQ was defined as the lowest concentration where the S/N was 10 : 1 for any product ion. The stabilities of CLE and CLE-d 9 were determined by analyzing bench top, processed, freeze – thaw and long-term samples. For each form of stability testing, three replicates were used and were compared with control samples at the same concentration that were extracted the day of LC – MS-MS analysis. Bench top samples were spiked with both analytes and stored at room temperature for at least 24 h before extraction was performed. Processed samples were spiked with CLE and CLE-d 9, extracted and reconstituted. However, they were then stored at 68C for 48 h before injection. Freeze – thaw samples were spiked and stored at 2208C for at least 48 h. The samples were then thawed to room temperature. This process was repeated two more times. The samples were then extracted and analyzed by the LC – MS-MS. Long-term samples were spiked and stored at 2208C for at least 90 days. The samples were then extracted and injected onto the LC – MS-MS. The abundance of the quantitation ion was used for both drugs in all experiments performed. Matrix effect, extraction recovery and process efficiency were all determined by spiking three blank urine samples with CLE and CLE-d 9 before the extraction process and three blank urine samples after the extraction process. Also, three neat standards were made at the same concentration. Matrix effect was calculated by dividing the mean abundance of the samples spiked after the extraction by the mean abundance of the neat standards, multiplied by 100 and then 100 was subtracted. If the number is positive, then there is enhancement and if it is negative, then there is suppression. Extraction recovery was determined by dividing the mean abundance of the samples supplemented before the extraction by the mean abundance of the samples supplemented after extraction and multiplied by 100. Process efficiency was evaluated by dividing the mean abundance of the samples supplemented before the extraction by the mean abundance of the neat standards and multiplied by 100. Accuracy and precision were calculated using 30 replicates of both high and low controls. Six samples at each concentration were run on five different days. The intra-, interday and total Confirmation and Quantification of Clenbuterol 131
precision were then calculated, along with the accuracy, using the ANOVA approach (10).
Results and Discussion Method validation The method for quantifying CLE in equine urine was completely and successfully validated, according to the Scientific Working Group for Forensic Toxicology recommendations (11). Blank urine extractions were found not to contain any traces of CLE and there was also no carry over or contamination observed. LOD was determined to be 50 pg/mL and the LLOQ was 100 pg/mL. The stability findings for CLE and CLE-d 9 are presented in Supplementary data, Table S1. For freeze–thaw, bench top and processed testing, degradation was shown to be ,15%. Long-term stability had more degradation for both CLE and CLE-d 9 (28 and 33%, respectively). Matrix effect, extraction recovery and process efficiency were calculated using ion suppression testing. The results for CLE were 288% for matrix effect, 100% for extraction recovery and 12% for process efficiency. The results for CLE-d 9 were 278, 63 and 14%, respectively. Five separate calibration curves were prepared with a total of 30 replicates of both high-and low-quality control urine preparations. Each curve was linear and had a correlation coefficient of .0.98. The representative calibration curve is presented in Supplementary data, Figure S1. Accuracy and precision were calculated using the data from the quality control samples. Intraand interday precision were found to be between 2 and 8% for high- and low-quality control samples, while total precision was 6 and 8%, respectively. Accuracy was calculated to be 2 and 14%, respectively. As this method was intended to be used to quantify CLE above a threshold, expanded uncertainty was calculated using the simplified Guide to the Expression of Uncertainty in Measurement (10). The calculated uncertainty at a 95% confidence level was 11% across the range.
Controlled administration samples The concentrations of CLE in the pooled urine samples from the two horses are presented in Supplementary data, Table S2. A time versus concentration curve was prepared using these data (Supplementary data, Figure S2). As expected, there was no CLE in the predose samples, but its concentration immediately increased in the first postdose sample after 2 h. The concentration of CLE peaked after 8 h and began to decline. The drug was still present in the sample collected 48 h postadministration, indicating that CLE is detected in urine for extended periods of time. This is consistent with previous reports of CLE being detected between 21 and 28 days postadministration (8).
Pilot program During the pilot program that was put into place, 50 of the 577 samples tested had detectable CLE in them, while only 1 was detected by ELISA. Only 12 of the 50 samples, 6 harness and 6 thoroughbred, were confirmed with the same extraction to be over the threshold of 140 pg/mL. Thus, 2% of the samples in the pilot program would have been reported as CLE overages.
132 Bishop et al.
Conclusion A new method to identify and quantify CLE was developed and validated in order to comply with the proposed threshold set forth by the RMTC and adopted by the Illinois Racing Board. A LC–MS-QQQ method substantially increases the sensitivity compared with previously utilized ELISA screening. The method validation study demonstrated acceptable stability, precision and accuracy. This method was used to quantify CLE in samples obtained from an administration study and from a pilot program of the Illinois Racing Board. A total of 2% of horses were found to contain CLE over the proposed threshold.
Supplementary data Supplementary data are available at Journal of Analytical Toxicology online.
References 1. Sasse, H.H., Hajer, R. (1978) NAB 365, a b2 receptor sympathomimetic agent: clinical experience in horses with lung disease. Journal of Veterinary Pharmacology and Therapeutics, 1, 241– 243. 2. Kearns, C.F., McKeever, K.H., Malinowski, K., Struck, M.B., Takashi, A. (2001) Chronic administration of therapeutic levels of clenbuterol acts as a repartitioning agent. Journal of Applied Physiology, 91, 2064–2070. 3. Erichsen, D.F., Aviad, A.D., Schultz, R.H., Kennedy, T.J. (1994) Clinical efficacy and safety of clenbuterol HCl when administered to effect in horses with chronic obstructive pulmonary disease (COPD). Equine Veterinary Journal, 26, 331–336. 4. Association of Racing Commissioners International, Inc. Uniform Classification Guidelines for Foreign Substances and Recommended Penalties and Model Rule. (2014), pp. ii – vii. http://arcicom. businesscatalyst.com/assets/uniform-classification-guidelines% 2c-version-07–00.pdf (accessed 5, Aug 2014). 5. Racing Medication and Testing Consortium. RMTC Controlled Therapeutic Medications List. (2014). http://www.rmtcnet.com/ content_landing_withdrawal.asp (accessed 4, Aug 2014). 6. Harkins, J.D., Robinson, N.E., Woods, W.E., Lehner, A.F., Smith, M.D., Gates, R.S. et al. (2000) Intratracheal clenbuterol in the horse: its pharmacological efficacy and analytical detection. Journal of Veterinary Pharmacology and Therapeutics, 23, 251–260. 7. Soma, L.R., Uboh, C.E., Guan, F., Moate, P., Luo, Y., Teleis, D. et al. (2004) Pharmacokinetics and disposition of clenbuterol in the horse. Journal of Veterinary Pharmacology and Therapeutics, 27, 71 –77. 8. Knych, H.K., Mitchell, M.M., Steinmetz, S.J., McKemie, D.S. (2014) Detection, pharmacokinetics and cardiac effects following administration of clenbuterol to exercised horses. Equine Veterinary Journal, 46, 380– 385. 9. Harkins, J.D., Woods, W.E., Lehner, A.F., Fisher, M., Tobin, T. (2001) Clenbuterol in the horse: urinary concentrations determined by ELISA and GC/MS after clinical doses. Journal of Veterinary Pharmacology and Therapeutics, 24, 7 –14. 10. LeBeau, M.A. (2011) Measuring and reporting uncertainty. In Watts, J., Moffat, A.C., Osselton, M.D., Widdop, B. (eds). Clarke’s Analysis of Drugs and Poisons, Vol. 1, 4th edition. Pharmaceutical Press, Gurnee, IL, pp. 371– 389. Chapter 23. 11. Scientific Working Group for Forensic Toxicology (SWGTOX). Standard Practices for Method Validation in Forensic Toxicology. 2013. http://www.swgtox.org/documents/Validation3.pdf (accessed Aug 14, 2014).
Article
Confirmation and Quantification of Clenbuterol in Horse Urine Using Liquid Chromatography Tandem Mass Spectrometry Triple Quadrupole Jennifer Bishop*, Brendan Heffron, Lisa Taddei, Marc Benoit, Laura Hurt, Sara Costello, Melissa Gross and Adam Negrusz Animal Forensic Toxicology Laboratory, Department of Biopharmaceutical Sciences, College of Pharmacy, University of Illinois at Chicago, 2242 West Harrison Street, Chicago, IL 60612, USA *Author to whom correspondence should be addressed. Email: [email protected]
Clenbuterol (CLE) is used in horses as a bronchodilator and for its anabolic steroid-like effects. CLE is a Class 3 drug according to current Association of Racing Commissioners International (ARCI) Uniform Classification Guidelines. The Racing Medication and Testing Consortium recommended a urine CLE threshold of 140 pg/mL after careful scientific review of the results of studies describing the disposition of CLE in the horse and this threshold was adopted by the ARCI. Enzyme-linked immunosorbent assay was previously used to screen samples for CLE in Illinois, but could not detect such low concentrations in urine. Thus, a liquid –liquid extraction of CLE from urine followed by quantification by liquid chromatography – tandem mass spectrometry was developed and validated. Method validation included testing stability, ion suppression and enhancement, precision, accuracy and uncertainty. Intra-, interday and total precision and accuracy were calculated for each control and found to be within the + 15% acceptance range. The Guide to the Expression of Uncertainty in Measurement approach was used to calculate uncertainty, which was 11% at the 95% confidence level. In the past 5 years, only 15 samples were reported as positive for CLE in Illinois. This new method was used in a pilot program to screen and confirm samples received from thoroughbred and harness horses.
Introduction Clenbuterol (CLE) is a bronchodilator and is the only b2 agonist approved by the Federal Drug Administration for use in horses. However, CLE can alter performance in race horses, including control of bronchospasms and anabolic effects. Horses that have been given chronic administration of CLE have shown significant decreases in percent fat and fat mass, and increases in fatfree mass (1 –3). VentipulminTM syrup containing CLE hydrochloride is approved by the US Food and Drug Administration for use in horses to treat inflammatory airway disease and chronic obstructive pulmonary disease. Anecdotal reports suggest that some trainers may administer CLE in various formulations in an attempt to achieve an anabolic effect particularly since the introduction of rules of racing that prohibit the use of anabolic steroids. CLE is a Class 3 drug according to current Association of Racing Commissioners International Uniform Classification Guidelines (4). Due to the fact that CLE is used therapeutically, the Racing Medication and Testing Consortium (RMTC) suggested a threshold, which has been accepted by the Illinois Racing Board (5). Previous research showed that the lowest effective dose of CLE is 0.8 mg/kg. Multiple methods of administration have been tested for efficacy, and the most effective ones were oral and intravenous (i.v.) administration (1). CLE given by intratracheal injection
showed no effect on horses that had decreased lung function, and therefore, is not considered an effective administration of the drug (6). Although oral administration was shown to have an effect for 1 h longer than i.v. administration, there is no pharmacokinetic difference between the two routes of administrations (1, 7). After 22 horses were given an oral dose of 0.8 mg/kg bwt twice a day for a total of 30 days, the highest concentration of CLE in urine found 14 days postadministration was 140 pg/mL (8) and it was chosen by the RMTC as the recommended threshold. Enzyme-linked immunosorbent assay (ELISA) was previously used as the screening method for CLE in Illinois based on local decisions. Even though the manufacturer’s limit of detection (LOD) is 1 ng/mL and previous reports have shown the LOD to be 0.1 ng/mL, the LOD of the University of Illinois at Chicago Animal Forensic Toxicology Laboratory was established to be 5 ng/mL (9). Thus, a method based on liquid –liquid extraction of CLE and its internal standard from urine followed by quantification by liquid chromatography mass spectrometry (LC – MS) was developed and validated (8). The primary objective of this study was to develop a method to confirm and quantify CLE in race horses. Due to the much lower screening levels, a pilot program was put into place for 3 months in order to determine how many horses would test positive over the new threshold and as a result, the violations would have been reported. Experimental Pilot program During the time before the threshold was implemented, a pilot program was performed in order to determine how many samples would have been reported over the threshold. A total of 577 samples (304 harness and 273 thoroughbred horses) were screened for CLE. This was only a small percentage of the total number of samples coming into the laboratory. CLE administration and sample collection The administration was provided by Interstate Drug Testing Alliance and Iowa State University. Two mare, quarter horses were used in this study. CLE hydrochloride was prepared by dissolving 2.59 mg in 5 mL of sterile saline. The solution was then filtered. The lowest therapeutic dose of 0.8 mg/mL was injected intravenously based on their weights of 643 and 606 kg. Predose urine was collected from both horses and samples were then collected at 2, 4, 8, 24 and 48 h postdose. Urine was collected via Foley catheter from both horses, and was pooled at each time collection.
# The Author 2014. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]
Reference compounds CLE hydrochloride and clenbuterol-d 9 (CLE-d 9) hydrochloride powders were obtained from Sigma-Aldrich (St. Louis, MO, USA). Stock solutions were made at 1 mg/mL in methanol for both CLE and the internal standard. Working solutions were made by diluting the stock solutions to a concentration of 1 mg/mL.
The parent ions that were used for detecting CLE and CLE-d 9 (CLE-d 9) were m/z 277.1 and 286.2, respectively. Product ions for CLE were 203, 168 and 132.1. The most abundant product ion was 203 and was therefore used for quantifying. CLE-d 9 had product ions of 204 and 66.2.
Method validation Reagents Methyl tert-butyl ether, methanol, formic acid, acetonitrile, water, glacial acetic acid and potassium carbonate were obtained from Fisher Scientific (Hanover Park, IL, USA), and were HPLC grade or better. Sodium acetate trihydrate was obtained from Mallinckrodt Pharmaceuticals (St. Louis, MO, USA). b-Glucuronidase was obtained from Campbell Science (Rockford, IL, USA). Blank equine urine was obtained from Bioreclamation LLC (Westbury, NY, USA).
Standard curve and quality control preparations Six curve points were made in blank equine urine at concentrations of 100, 200, 400, 800, 1600 and 3200 pg/mL. Quality control preparations were made at two concentrations in blank equine urine (2000 and 250 pg/mL).
Extraction method For all blank, quality control, curve points and samples, 2 mL of equine urine was placed in screw cap tubes. To each sample, 500 mL of 0.9 M sodium acetate buffer containing internal standard at a concentration of 500 pg/mL and 50 mL of bglucuronidase were added. The samples were mixed, capped loosely and incubated overnight at 378C. After cooling, 2 mL of 0.5 M potassium carbonate and 5 mL of methyl tert-butyl ether were added to each sample. The tubes were capped, mixed by rotorack for 10 min and centrifuged for 10 min at 3000 rpm. The solvent layer was then transferred to a clean tube and evaporated to dryness under nitrogen in a water bath at 378C. Two hundred and fifty microliters of 0.2% formic acid in water was added to each tube. The samples were vortexed and transferred to an HPLC autosampler deep well. Twenty microliters from each sample were then injected onto the LC –MS-MS.
Instrumentation An Agilent 1200 Series HPLC coupled with an Agilent 6410 Series Triple Quadrupole Mass Spectrometer (LC – MS-QQQ) in positive ionization mode was used to detect and quantify CLE. The column used was an Agilent Poroshell 120 EC-C 18 2.1 i.d.100 mm, 2.7 mm particle diameter. The pump flow was set to 500 mL/min and the column temperature was kept at 308C. Solvent A was 0.2% formic acid in water and Solvent B was 90 : 10 acetonitrile : water. The initial concentration of Solvent B was 10% and ramped to 45% over 2 min. Then the concentration was increased to 100% over 1 min and held for 2 min before re-equilibrating to 10%. Gas temperature was set to 3508C while the gas flow was at 11 L/min. Ultra-high-purity nitrogen was used for the drying, sheath and collision gases. The nebulizer pressure was at 60 psi and the capillary voltage was 3500 V.
Method validation included selectivity, carryover, LOD, LLOQ, stability, extraction recovery, matrix effect, process efficiency, accuracy and precision. Selectivity was determined by analyzing three blank equine urine samples from three different sources. These samples were extracted and injected onto the LC – MS-MS to analyze for the presence of CLE and CLE-d 9 product ions based on injected standard retention times. Carryover and contamination were assessed by analyzing three blank samples that were injected after the highest concentration curve point. LOD was determined by finding the lowest concentration where the signal to noise (S/N) of any product ion was 5 : 1. LLOQ was defined as the lowest concentration where the S/N was 10 : 1 for any product ion. The stabilities of CLE and CLE-d 9 were determined by analyzing bench top, processed, freeze – thaw and long-term samples. For each form of stability testing, three replicates were used and were compared with control samples at the same concentration that were extracted the day of LC – MS-MS analysis. Bench top samples were spiked with both analytes and stored at room temperature for at least 24 h before extraction was performed. Processed samples were spiked with CLE and CLE-d 9, extracted and reconstituted. However, they were then stored at 68C for 48 h before injection. Freeze – thaw samples were spiked and stored at 2208C for at least 48 h. The samples were then thawed to room temperature. This process was repeated two more times. The samples were then extracted and analyzed by the LC – MS-MS. Long-term samples were spiked and stored at 2208C for at least 90 days. The samples were then extracted and injected onto the LC – MS-MS. The abundance of the quantitation ion was used for both drugs in all experiments performed. Matrix effect, extraction recovery and process efficiency were all determined by spiking three blank urine samples with CLE and CLE-d 9 before the extraction process and three blank urine samples after the extraction process. Also, three neat standards were made at the same concentration. Matrix effect was calculated by dividing the mean abundance of the samples spiked after the extraction by the mean abundance of the neat standards, multiplied by 100 and then 100 was subtracted. If the number is positive, then there is enhancement and if it is negative, then there is suppression. Extraction recovery was determined by dividing the mean abundance of the samples supplemented before the extraction by the mean abundance of the samples supplemented after extraction and multiplied by 100. Process efficiency was evaluated by dividing the mean abundance of the samples supplemented before the extraction by the mean abundance of the neat standards and multiplied by 100. Accuracy and precision were calculated using 30 replicates of both high and low controls. Six samples at each concentration were run on five different days. The intra-, interday and total Confirmation and Quantification of Clenbuterol 131
precision were then calculated, along with the accuracy, using the ANOVA approach (10).
Results and Discussion Method validation The method for quantifying CLE in equine urine was completely and successfully validated, according to the Scientific Working Group for Forensic Toxicology recommendations (11). Blank urine extractions were found not to contain any traces of CLE and there was also no carry over or contamination observed. LOD was determined to be 50 pg/mL and the LLOQ was 100 pg/mL. The stability findings for CLE and CLE-d 9 are presented in Supplementary data, Table S1. For freeze–thaw, bench top and processed testing, degradation was shown to be ,15%. Long-term stability had more degradation for both CLE and CLE-d 9 (28 and 33%, respectively). Matrix effect, extraction recovery and process efficiency were calculated using ion suppression testing. The results for CLE were 288% for matrix effect, 100% for extraction recovery and 12% for process efficiency. The results for CLE-d 9 were 278, 63 and 14%, respectively. Five separate calibration curves were prepared with a total of 30 replicates of both high-and low-quality control urine preparations. Each curve was linear and had a correlation coefficient of .0.98. The representative calibration curve is presented in Supplementary data, Figure S1. Accuracy and precision were calculated using the data from the quality control samples. Intraand interday precision were found to be between 2 and 8% for high- and low-quality control samples, while total precision was 6 and 8%, respectively. Accuracy was calculated to be 2 and 14%, respectively. As this method was intended to be used to quantify CLE above a threshold, expanded uncertainty was calculated using the simplified Guide to the Expression of Uncertainty in Measurement (10). The calculated uncertainty at a 95% confidence level was 11% across the range.
Controlled administration samples The concentrations of CLE in the pooled urine samples from the two horses are presented in Supplementary data, Table S2. A time versus concentration curve was prepared using these data (Supplementary data, Figure S2). As expected, there was no CLE in the predose samples, but its concentration immediately increased in the first postdose sample after 2 h. The concentration of CLE peaked after 8 h and began to decline. The drug was still present in the sample collected 48 h postadministration, indicating that CLE is detected in urine for extended periods of time. This is consistent with previous reports of CLE being detected between 21 and 28 days postadministration (8).
Pilot program During the pilot program that was put into place, 50 of the 577 samples tested had detectable CLE in them, while only 1 was detected by ELISA. Only 12 of the 50 samples, 6 harness and 6 thoroughbred, were confirmed with the same extraction to be over the threshold of 140 pg/mL. Thus, 2% of the samples in the pilot program would have been reported as CLE overages.
132 Bishop et al.
Conclusion A new method to identify and quantify CLE was developed and validated in order to comply with the proposed threshold set forth by the RMTC and adopted by the Illinois Racing Board. A LC–MS-QQQ method substantially increases the sensitivity compared with previously utilized ELISA screening. The method validation study demonstrated acceptable stability, precision and accuracy. This method was used to quantify CLE in samples obtained from an administration study and from a pilot program of the Illinois Racing Board. A total of 2% of horses were found to contain CLE over the proposed threshold.
Supplementary data Supplementary data are available at Journal of Analytical Toxicology online.
References 1. Sasse, H.H., Hajer, R. (1978) NAB 365, a b2 receptor sympathomimetic agent: clinical experience in horses with lung disease. Journal of Veterinary Pharmacology and Therapeutics, 1, 241– 243. 2. Kearns, C.F., McKeever, K.H., Malinowski, K., Struck, M.B., Takashi, A. (2001) Chronic administration of therapeutic levels of clenbuterol acts as a repartitioning agent. Journal of Applied Physiology, 91, 2064–2070. 3. Erichsen, D.F., Aviad, A.D., Schultz, R.H., Kennedy, T.J. (1994) Clinical efficacy and safety of clenbuterol HCl when administered to effect in horses with chronic obstructive pulmonary disease (COPD). Equine Veterinary Journal, 26, 331–336. 4. Association of Racing Commissioners International, Inc. Uniform Classification Guidelines for Foreign Substances and Recommended Penalties and Model Rule. (2014), pp. ii – vii. http://arcicom. businesscatalyst.com/assets/uniform-classification-guidelines% 2c-version-07–00.pdf (accessed 5, Aug 2014). 5. Racing Medication and Testing Consortium. RMTC Controlled Therapeutic Medications List. (2014). http://www.rmtcnet.com/ content_landing_withdrawal.asp (accessed 4, Aug 2014). 6. Harkins, J.D., Robinson, N.E., Woods, W.E., Lehner, A.F., Smith, M.D., Gates, R.S. et al. (2000) Intratracheal clenbuterol in the horse: its pharmacological efficacy and analytical detection. Journal of Veterinary Pharmacology and Therapeutics, 23, 251–260. 7. Soma, L.R., Uboh, C.E., Guan, F., Moate, P., Luo, Y., Teleis, D. et al. (2004) Pharmacokinetics and disposition of clenbuterol in the horse. Journal of Veterinary Pharmacology and Therapeutics, 27, 71 –77. 8. Knych, H.K., Mitchell, M.M., Steinmetz, S.J., McKemie, D.S. (2014) Detection, pharmacokinetics and cardiac effects following administration of clenbuterol to exercised horses. Equine Veterinary Journal, 46, 380– 385. 9. Harkins, J.D., Woods, W.E., Lehner, A.F., Fisher, M., Tobin, T. (2001) Clenbuterol in the horse: urinary concentrations determined by ELISA and GC/MS after clinical doses. Journal of Veterinary Pharmacology and Therapeutics, 24, 7 –14. 10. LeBeau, M.A. (2011) Measuring and reporting uncertainty. In Watts, J., Moffat, A.C., Osselton, M.D., Widdop, B. (eds). Clarke’s Analysis of Drugs and Poisons, Vol. 1, 4th edition. Pharmaceutical Press, Gurnee, IL, pp. 371– 389. Chapter 23. 11. Scientific Working Group for Forensic Toxicology (SWGTOX). Standard Practices for Method Validation in Forensic Toxicology. 2013. http://www.swgtox.org/documents/Validation3.pdf (accessed Aug 14, 2014).
