A Study of Non-steroidal Anti-inflammatory Drugs in the Urine of the Racing Greyhound

The purpose of this study was to develop a screening procedure for the detection of non-steroidal anti-inflammatory drugs (NSAIDs) in the urine of the racing greyhound. The study was divided into three parts, dealing with the evaluation and selection of an appropriate analytical methodology, the investigation of acidic endogenous components of greyhound urine and the analysis of selected NSAIDs in urine following administration at therapeutic levels to greyhounds. (1) in the first part of the study, the analytical methods examined were isocratic high pressure liquid chromatography (HPLC) with ultraviolet (UV) detection, capillary gas chromatography (GC) with temperature programming and flame ionisation detection and gas chromatography-mass spectrometry (GC-MS). Twelve NSAIDs in common use in the UK were selected as test compounds for the evaluation of the separating ability, selectivity and sensitivity of detection of the three methods. (a) HPLC : For one compound, phenylbutazone, the HPLC capacity factor (k' ) was measured in seven reversed-phase solvent systems using an octadecylsilane column. The remaining compounds were examined on two selected systems. In addition, fluorescence spectrometry and electrochemical detection (ECD) were examined to assess their value as additional HPLC detection systems which might increase the selectivity or sensitivity of the method. All of the test compounds were eluted from the column and could be detected by UV spectrometry at a sample size of 5 nanograms on-column. Five of the test compounds could be detected by fluorescence spectrometry and seven by ECD. However, none of the solvent systems examined resolved all of the test compounds from each other. (b) GLC: Three chemical derivatives were evaluated to increase the thermal stability and improve the chromatographic behaviour of the test compounds - the methyl, trimethylsilyl and tertiary-butyldimethylsilyl derivatives. The retention indices of the three derivatives for each substance were recorded using a glass capillary column coated with dimethylsilicone (non-polar) stationary phase. Three compounds, all pyrazolidine diones, could be chromatographed without derivatisation but produced a mixture of products following reaction with the methylating agent, diazomethane. These were further examined by mass spectrometry. (c) GC-MS : Gas chromatography with mass spectrometric detection was carried out using similar GC conditions to those described above. Electron impact spectra were recorded at 70 ev for each of the derivatives of the test substances and possible fragmentation reactions giving rise to the most prominent ions in the spectra were proposed. The reaction products of pyrazolidinedione NSAIDs produced by reaction with diazomethane were characterised by GC-MS and tentative structures were proposed in which methylene addition had occurred at the enol-oxygen atom and at the carbonyl group, to give enol ethers and oxiranes respectively. It was concluded that GC-MS was the only method which would have sufficient selectivity and separating ability for use in a screening procedure and this method was used in the rest of the study. (2) The second part of the study established the normal pattern of organic acids present in greyhound urine, which would be co-extracted with any NSAIDs present and might lead to interference in the detection and quantitation of the drugs. (3) In the third part of the study five test compounds (ibuprofen, naproxen, ketoprofen, mefenamic acid and phenylbutazone) were administered at therapeutic levels to greyhounds. Blood and urine samples were collected serially in the ibuprofen and mefenamic acid experiments. Urine samples only were collected for the remaining three compounds. urinary creatinine concentrations were measured for each sample. The urinary components were extracted as before with XAD-2 resin and glucuronide and sulphate conjugates present in the samples were hydrolysed by incubation with helix-pomatia extract. Plasma samples were extracted using heptane:ethyl acetate. All samples were methylated using diazomethane. The five test compounds were detected and quantified in the urine and plasma samples collected. Plasma drug concentration/time curves and excretion profiles were obtained for ibuprofen and mefenamic acid. Metabolites of ibuprofen and naproxen were observed in urine samples. The importance of parallel creatinine measurements to correct for variation in the urinary volume was shown in the naproxen study. The implications of the results with respect to the detection of these five drugs and the interpretation of their concentrations during a screening procedure at a race meeting are discussed