UC Davis

The horseracing industry has strict regulations regarding the presence of chemical substances in equine athletes at the time of competition. Metformin, a commonly prescribed diabetes medication in human medicine, is considered a prohibited substance in horseracing, as there is no accepted therapeutic use for this compound in young, fit horses. Recent metformin detections in post-race blood and urine samples resulting in adverse analytical findings and severe disciplinary action against the trainers of the horses involved necessitates further study of this drug in horses. This thesis aims to characterize the pharmacokinetic behavior of metformin in the horse and explore environmental contamination as a means of exposure to inform regulation of metformin in horseracing.In the first study, a pharmacokinetic (PK) study of metformin was performed utilizing twelve horses. Metformin was administered orally and intravenously in a balanced, randomized, two-way crossover design. Serum drug concentrations were determined using a highly sensitive liquid chromatography-tandem mass spectrometry (LC-MS/MS) assay and data was analyzed via noncompartmental analysis (NCA) for calculation of select pharmacokinetic parameters. Subsequently, concentrations were analyzed using a non-linear mixed effects (NLME) population model. The final model was then used to perform Monte Carlo simulations utilizing the estimated fixed effects and interindividual variability of the study horses to generate serum concentrations in a simulated population of 1000 Thoroughbred horses. These simulations allowed for determination of a detection time using the current screening limit and estimation of withdrawal times for metformin corresponding to several additional proposed screening limits. The current serum screening limit of 0.5 ng/mL necessitated a prolonged detection time of 475 hours (~20 days).To better understand the prolonged detection time of metformin in horses, a second PK study was performed wherein six horses were administered metformin orally (15 grams) for evaluation of sequestration into erythrocytes, a phenomenon that has been reported in humans. Metformin concentrations were determined in blood matrices, including plasma, serum, hemolyzed plasma, hemolyzed serum, serum from a serum separator tube (SST), red blood cells (RBCs), and whole blood, collected post administration. The ratios of blood to plasma average metformin concentrations were found to be >10 at later time points. The average concentrations in serum and plasma were also compared and found to be >15% different at several time points. The results of this study suggest RBCs act as a reservoir for metformin, contributing to extended detection time in blood samples.Another study was performed to examine environmental exposure as a source for metformin detection in blood and urine of horses. Metformin is widely prescribed in human medicine and is excreted in urine as unchanged metformin. To simulate proposed metformin exposure at racetracks, six horses were exposed to metformin-contaminated urine. Blood and urine samples were collected for determination of metformin concentration following exposure. The blood matrices analyzed included plasma, serum, serum collected in serum separator tubes, whole blood, and RBCs. Over the course of the study all horses had detectable amounts of metformin in at least one matrix for at least one time point; urine concentrations were much greater than blood. The results of this study indicate that environmental contamination may be a possible route of metformin exposure for horses housed at racetracks.