5h.1 Background and Evidence Review
DM is a known risk factor for urinary tract infections (UTIs), and this may be attributable to glycosuria enhancing bacterial growth in the urinary tract or to bladder dysfunction impairing complete bladder emptying (77, 78). The mechanism of action of SGLT 2 inhibitors suggests a theoretical increased risk of UTIs through the enhancement of glycosuria. In 2015, the FDA reported 19 cases of urosepsis and pyelonephritis in individuals on SGLT 2 inhibitor therapy between March 2013 and October 2014, and subsequently issued a warning surrounding the risk of UTIs with SGLT 2 inhibitor use (18).
Except for VERTIS CV and EMPEROR-PRESERVED, this finding is not reflected in the randomized data from the large placebo-controlled trials (1, 2, 7, 13, 14, 21, 23, 24, 35, 37, 45). It is notable that VERTIS CV did report a statistically significant increase in any UTIs in the ertugliflozin arm (absolute increase in risk with ertugliflozin 5mg vs placebo +2.1% over 3.5 years, 95%CI 0.4-3.7%; 15mg vs placebo, +1.8%, 95%CI 0.2-3.5%), but found no difference in the subset of these infections which were serious (<10% of UTIs were serious). In EMPEROR-PRESERVED, absolute UTI risk was increased by ~0.8% per year (3, 14).
Meta-analysis of the 13 large randomized trials in SGLT 2 inhibition demonstrate a small excess risk for UTI (RR 1.08, 95%CI 1.02-1.15), contributing to about 1 in 14 UTIs experienced in a population treated with SGLT 2 inhibitors (3). Meta-analysis limited only to serious UTI did not show a significant increase in risk (RR 1.07, 95%CI 0.90-1.27).
The lack of serious UTI risk with SGLT 2 inhibition might be explained by the hypothesis that any effects of glycosuria on potentiating bacterial growth are countered by those of diuresis and polyuria that prevent bacterial ascension of the urinary tract (79). Alternatively, glycosuria per se is actually not a common precipitant for UTIs.
In CKD, there is reassuring data (80). In CREDENCE, the event rate for UTI per 1000 person-years was 48.3 in the canagliflozin-treated group versus 45.1 in the placebo arm (HR=1.08, 0.90–1.29), and in SCORED, the event rate for UTI per 1000 person years was 86 in the sotagliflozin-treated group versus 83 in those allocated placebo (HR=1.04, 0.94-1.16) (7, 24). Rates of serious UTI in EMPA-KIDNEY were similar in empagliflozin and placebo groups (0.81 vs 0.84 events per 100 person years, respectively) (1).
5h.2 Recommendations for Implementations
- We recommend temporary discontinuation of SGLT 2 inhibitors when treating acute pyelonephritis or urosepsis (see sick-day guidance 5a.1.2) (Grade 1C).
Rationale: Randomized data from major trials show the increased risk of UTIs with SGLT 2 inhibitors is small. However, these drugs are being prescribed in people who have a high risk of UTIs and effective prompt management of these infections should be undertaken.