Metformin in the Horse: Pharmacokinetics and Detection Times Using Monte Carlo Simulations

The racehorse industry has strict regulations regarding the detection of prohibited substances in horses. Metformin, a diabetes medication, is a prohibited substance that has been reported in post-race blood and urine samples collected from racehorses. For further characterization of the disposition of metformin, 12 Thoroughbred horses were administered metformin orally and intravenously in a randomized, balanced, two-way crossover design. Serum and urine samples were collected, and drug concentrations determined via liquid chromatography-tandem mass spectrometry. The serum data were analyzed using both noncompartmental analysis and a population pharmacokinetic model. Metformin concentrations were below the LOQ (0.25 ng/mL) in six of 12 horses by Day 11 postadministration, and below the LOQ for all horses on Day 25. The maximum serum concentration of metformin (mean ± SD) was 941.0 ± 467.8 ng/mL, and the mean terminal t_1/2 was 85.8 ± 15.1 h. Based on Monte Carlo simulations the time that serum metformin concentrations fell below the proposed HISA Anti-Doping and Medication Control (ADMC) minimum reporting level (MRL; 0.5 ng/mL) in a simulated population of 1000 Thoroughbred horses was 475 hrs (~20 days). Metformin concentrations in urine fluctuated significantly between and within individual horses, and there was not a consistent relationship between blood and urine samples across time points. Results of the present study demonstrate a prolonged detection time; thus, a prolonged withdrawal time is needed to prevent a positive finding following exposure to metformin. Additionally, these results suggest that blood is the preferred matrix for regulatory purposes due to the inconsistent elimination in urine.