PWI-002 - SHORT-TERM PHARMACODYNAMIC EFFECTS OF CANAGLIFLOZIN IN HEALTHY AMISH VOLUNTEERS

S. Bargal¹, M. Montasser², E. Streeten², H. Whitlatch², J. O'Connell², Z. Shahidzadeh Yazdi², S. Taylor², A. Beitelshees³; ¹University of Maryland School of Medicine, Baltimore, MD, USA, ²University of Maryland School of Medicine, University of Maryland School of Medicine, USA, ³University of Maryland, Baltimore, MD, USA.

Background: Canagliflozin is a sodium-glucose cotransporter-2 inhibitor (SGLT2-I), a class of antidiabetics that has been recently considered as a first choice for many patients with type 2 diabetes and comorbidities due to their cardioprotective and renoprotective effects in addition to HbA1c-lowering, weight loss, and blood pressure-lowering. Nevertheless, SGLT2-Is also have significant risks - including bone loss/fractures (canagliflozin), urosepsis, and ketoacidosis. There is significant variability in response to these drugs and therefore our goal involves conducting a GWAS to identify genetic variants associated with both efficacy and safety endpoints. In this abstract, we focus on reporting our preliminary pharmacodynamic findings of this Genetics of Response to Canagliflozin (GRC) study.

Methods: We have completed a short-term intervention in 403 healthy volunteers from the Old Order Amish population. Our study enrolled healthy volunteers as a first step in the discovery process to maximize the signal-to-noise ratio by minimizing confounders present in diabetic patients. Participants were administered canagliflozin 300 mg/day for 5 days, completed a 24-hour urine collection, and had blood drawn at baseline, day 3, and day 6. Glucosuria was the primary endpoint. Secondary endpoints included additional biomarkers for efficacy and safety related to bone, cardiovascular disease, and ketosis.

Results: 403 participants (173 men and 230 women) completed the study with average age 48 ± 14 years. Canagliflozin induced an average glucosuria of 34.4 ± 9.5 ggluc/gcr. The significant covariates associated with glucosuria in a linear regression model were sex (p=0.00357), fasting plasma glucose (p=0.00210) and eGFR (p=6.28x10-15). When participant data were binned into tertiles, within each tertile there was still substantial variation observed in glucosuria. Change in secondary endpoints are shown in the Figure.

Conclusion: Preliminary pharmacodynamic data emphasize that there is substantial unexplained variability in glucosuria. This GRC study, therefore, offers the opportunity to identify genetic predictors of safety and efficacy biomarkers for SGLT2-I response that could help explain some of this variation and hence help guide diabetes treatment decisions in the future.