Canagliflozin and risk of genital infections and urinary tract infections in people with diabetes mellitus and kidney disease- a post-hoc analysis of the credence trial

Amy Kang,Brendon Neuen,Hiddo Lambers Heerspink,Gian Luca Di Tanna,Bruce Neal,Hong Zhang,Carinna Hockham,Rajiv Agarwal,George Bakris,David M. Charytan,Dick De Zeeuw,Tom Greene,Adeera Levin,Carol Pollock,David Wheeler,Bernard Zinman,Kenneth W. Mahaffey,Vlado Perkovic,Meg Jardine

DOI: https://doi.org/10.1093/ndt/gfaa142.p1013

IF: 7.1857

2020-01-01

Nephrology Dialysis Transplantation

Abstract:Abstract Background and Aims To describe genital mycotic infections (GMI) and urinary tract infections (UTI) in the CREDENCE trial, determine whether canagliflozin increased the risk of these infections overall and in subgroups, and describe predictors of risk for genital mycotic infections. Method The CREDENCE trial randomised people with type 2 diabetes and albuminuric stage 2 and 3 chronic kidney disease to canagliflozin 100mg daily or placebo. We analysed the risk of GMI and UTI with canagliflozin compared to placebo overall and in patient subgroups. The primary analysis was conducted in the on-treatment population, as the more conservative approach with sensitivity analyses conducted using an intention-to-treat population. When canagliflozin increased risk, we determined patient risk factors for GMIs using multivariable Cox regression models adjusting for age, gender, race, markers of disease severity (body mass index (BMI), haemoglobin A1c, diabetes duration, other glucose lowering medications at baseline and kidney function). Results Overall 31/2905 (1.1%) men and 32/1492 (2.1%) women experienced 91 GMIs and 166/2905 (5.7%) men and 300/1492 (20.1%) women experienced 669 UTIs. Canagliflozin increased the risk of GMI (HR 3.83 [95% CI 2.08-7.06] p<0.0001). The hazard ratio for canagliflozin compared to placebo was consistent across most subgroups, though the risk with canagliflozin was greater in those with a higher BMI (HR 5.91 [95% CI 2.65-13.15] for BMI ≥30 kg/m2 vs HR 1.36 [95% CI 0.47-3.92] for BMI<30 kg/m2, p interaction=0.03) and was higher in men (HR 9.30 [95% CI 2.83-30.60] vs HR 2.10 [95% CI 1.00-4.45] for men and women respectively, p interaction=0.04). In those who were randomised to canagliflozin, independent risk factors for GMI were higher BMI (HR 1.53 [95% CI 1.29-1.83] per 5 units p<0.0001) and longer diabetes duration (HR 1.18 [95% CI 1.01-1.40] per 5 years p=0.04). Canagliflozin did not affect the risk of UTI over placebo (HR 1.08 [95% CI 0.90-1.29] p=0.42) overall or in any subgroup, however risk was higher in women (HR 1.23 [95% CI 0.98-1.54] vs HR 0.82 [0.60-1.11] for women and men respectively, p interaction=0.04).58/669 (8.7%) UTIs but no GMIs were reported as serious. Drug was continued in 56/63 (89%) of first GMIs, with similar frequency of subsequent GMI in those continuing on canagliflozin (13/43, 30.2%) or placebo (4/13, 30.8%). Drug was continued in 385/466 (82.6%) first UTIs, with similar frequency of subsequent UTIs in those continuing on cangliflozin (50/199 (25.1%) or placebo 49/186 (26.3%). All findings were similar when conducted using an intention-to-treat approach. Conclusion Canagliflozin increased the risk of genital mycotic infections but not urinary tract infections. The risk of genital mycotic infections from canagliflozin over placebo was higher in men and those with higher BMI. In those treated with canagliflozin, higher BMI and longer diabetes duration independently predicted infection. Most participants continued treatment following their first infection with similar recurrence rates in the canagliflozin and placebo groups.These findings will be useful in clinical care, and help identify those at greatest risk for genital infections with canagliflozin treatment.