Comparative safety of different recommended doses of sodium-glucose cotransporter 2 inhibitors in patients with type 2 diabetes mellitus: a systematic review and network meta-analysis of randomized clinical trials.

Objective: The safety results of different recommended doses of sodium-glucose cotransporter 2 inhibitors (SGLT-2i) for patients with type 2 diabetes mellitus (T2DM) remain uncertain. This study aims to comprehensively estimate and rank the relative safety outcomes with different doses of SGLT-2i for T2DM. Methods: PubMed, Embase, the Cochrane Central Register of Controlled Trials, ClinicalTrials.gov, Chinese National Knowledge Infrastructure, WanFang database, and SinoMed database were searched from the inception to 31 May 2023. We included double-blind randomized controlled trials (RCTs) comparing SGLT-2i with placebo or another antihyperglycemic as oral monotherapy in the adults with a diagnosis of T2DM. Results: Twenty-five RCTs with 12,990 patients randomly assigned to 10 pharmacological interventions and placebo were included. Regarding genital infections (GI), all SGLT-2i, except for ertugliflozin and ipragliflozin, were associated with a higher risk of GI compared to placebo. Empagliflozin 10mg/d (88.2%, odds ratio [OR] 7.90, 95% credible interval [CrI] 3.39 to 22.08) may be the riskiest, followed by empagliflozin 25mg/d (83.4%, OR 7.22, 95%CrI 3.11 to 20.04)) and canagliflozin 300mg/d (70.8%, OR 5.33, 95%CrI 2.25 to 13.83) based on probability rankings. Additionally, dapagliflozin 10mg/d ranked highest for urinary tract infections (UTI, OR 2.11, 95%CrI 1.20 to 3.79, 87.2%), renal impairment (80.7%), and nasopharyngitis (81.6%) when compared to placebo and other treatments. No increased risk of harm was observed with different doses of SGLT-2i regarding hypoglycemia, acute kidney injury, diabetic ketoacidosis, or fracture. Further subgroup analysis by gender revealed no significantly increased risk of UTI. Dapagliflozin 10mg/d (91.9%) and canagliflozin 300mg/d (88.8%) ranked first in the female and male subgroups, respectively, according to the probability rankings for GI. Conclusion: Current evidence indicated that SGLT-2i did not significantly increase the risk of harm when comparing different doses, except for dapagliflozin 10mg/d, which showed an increased risk of UTI and may be associated with a higher risk of renal impairment and nasopharyngitis. Additionally, compared with placebo and metformin, the risk of GI was notably elevated for empagliflozin 10mg/d, canagliflozin 300mg/d, and dapagliflozin 10mg/d. However, it is important to note that further well-designed RCTs with larger sample sizes are necessary to verify and optimize the current body of evidence. Systematic Review Registration: https://www.crd.york.ac.uk/PROSPERO/, identifier CRD42023396023.