LAUSR

Background: We aimed to assess the safety and efficiency of the novel sodium glucose co-transporter 2 (SGLT2) inhibitor in combinations with insulin for type 1 and type 2 diabetes mellitus (T1DM and T2DM). Methods: We searched Medline, Pubmed, Embase, and the Cochrane Collaboration Library from January 2010 to December 2016 without restriction of language. FDA data and Clinical Trials (http://www.clinicaltrials.gov) were also searched. Study selection, data extraction, and evaluation of risk of bias were performed by 2 persons independently. The risk of bias was assessed by Cochrance System Evaluate Method and Q test was used to evaluate the heterogeneity between studies. We used random effect model to analyze the results by Revman 5.3. This meta-analysis has been registered at online public registry PROSPERO (registration number is: CRD42017054718). Results: Nine trials including 3069 patients were analyzed. Compared with control group, SGLT2 inhibitor produced absolute reduction in glycosylated hemoglobin A1c (HbA1c) (MD −1.35%, 95% confidence interval [CI] [−2.36 to −0.34], P = .009), fasting plasma glucose (FPG) (MD −1.01 mmol/L, 95%CI [−1.98 to 0.04], P = .04), insulin dosage (MD −4.85 U/24 hours, 95%CI [−7.42 to −2.29], P = .002), and body weight (MD −2.30 kg, 95%CI [−3.09 to −1.50], P < .00001). But the risk of hypoglycemia (OR 1.18, 95%CI [0.86, 1.61], P = . 30) and urinary tract infection (UTI) (OR 1.34, 95%CI [0.79, 2.27], P = .28) were proved as no difference and genital tract infection (GTI) with SGLT2 inhibitors was higher than control group (OR 2.96, 95%CI [1.05, 8.37], P = .04), in which cases were mild and responded to the therapy. According to the subgroup analysis, SGLT2 inhibitors had a similar effect in effective factors of both T1DM and T2DM, but the risk of GTI mainly increased in T2DM versus T1DM (T1DM OR 0.27 [0.01, 7.19], P = .43 vs T2DM OR 4.28 [2.00, 9.16], P = .0002). Conclusion: SGLT2 inhibitors have improved the HbA1c, FPG, and body weight when combined with insulin and decreased the dose of insulin without increasing the risk of hypoglycemia. However, SGLT2 inhibitor was proved to be related to the events of GTI, despite SGLT2 inhibitors appeared to be well tolerated. We suggest that more monitoring should be done to prevent the events of GTI, and more randomized controlled trials should be planned next step.