BackgroundA previous six-week (wk) study demonstrated the potential of the sodium-glucose linked transport inhibitor velagliflozin as a novel treatment for equine insulin dysregulation. The present study examined the safety and efficacy of velagliflozin over 16 wk. of treatment, and over 4 wk. of withdrawal. Twenty-four insulin dysregulated ponies were selected, based on their hyper-responsiveness to a diet challenge meal containing 3.8 g non-structural carbohydrates (NSC)/kg bodyweight (BW). Ponies with serum insulin > 90 μIU/mL either 2 or 4 h after feeding were enrolled, and randomly allocated to receive either velagliflozin (0.3 mg/kg BW orally once daily, n = 12), or a placebo (n = 10–12) for 16 wk. The subjects were fed 7.5 g NSC/kg BW/day to maintain a fat body condition. Safety was assessed through daily monitoring, veterinary examination, and the measurement of fasting blood glucose, biochemistry and haematology. Efficacy at reducing post-prandial hyperinsulinemia was assessed using a diet challenge every 8 wk. during treatment and 4 wk. after withdrawal.ResultsVelagliflozin was well accepted by all subjects and caused no adverse effects or hypoglycaemia. Post-prandial serum insulin (insulin Cmax) did not change significantly in the control animals over the entire study period (P = 0.101). In contrast, insulin Cmax (mean ± SE) concentrations fell over time in the velagliflozin-treated group from 205 ± 25 μIU/mL in wk. 0, to 119 ± 19 μIU/mL (P = 0.015) and 117 ± 15 μIU/ml (P = 0.029) after 8 and 16 wk. of treatment, respectively. Although the insulin Cmax in this group was not significantly lower than in controls at wk-8 (P = 0.061), it was lower at wk-16 (P = 0.003), and all 12 treated ponies were below the previously-determined risk threshold for laminitis at this time. After 4 wk. withdrawal, the insulin Cmax returned to 199 ± 36 μIU/mL in the treated group, with no rebound effect.ConclusionsVelagliflozin appears to be a promising and safe treatment for equine insulin dysregulation, bringing post-prandial insulin concentrations below the laminitis risk threshold, albeit without normalising them.