The efficacy and safety of dipeptidyl peptidase-4 inhibitors for type 2 diabetes: a Bayesian network meta-analysis of 58 randomized controlled trials

The aim is to evaluate the efficacy and safety of dipeptidyl peptidase-4 inhibitors (DPP4-I: sitagliptin, saxagliptin, linagliptin, vildagliptin and alogliptin) in patients with type 2 diabetes. We searched the Cochrane Library, PubMed, EMBASE, Chinese Biomedical Database (CBM), China National Knowledge Infrastructure (CNKI), and the Wanfang Database from inception to April, 2018. Randomized controlled trials were included if they compared the different versions of DPP4-I with each other or with placebo in treatment of type 2 diabetes. Bayesian network meta-analysis and pairwise meta-analysis were performed to evaluate the efficacy and safety of the different kinds of DPP4-I and placebo. The data were analyzed using STATA 12.0 and WinBUGS1.4 software. We identified 58 eligible studies (with 31356 patients) involving 14 treatment arms. Indirect comparison results showed that except for alogliptin, a decrease was found for all DPP4-I versus the placebo for hemoglobin A1c (HbA1c) with vildagliptin50 twice daily (BID) showing the highest probability. Linagliptin5 once daily (QD) decreased the level of fasting plasma glucose (FPG) the most for all DPP4-I versus the placebo; when comparing them with each other, alogliptin25QD was more effective when compared with sitagliptin100QD and vildaglipti50BID; linagliptin5qd had the highest decrease impact on body mass index (BMI). Except for hypoglycemia and upper respiratory tract infection (URTI), there are no statistical significance on incidence of adverse events and the body weight when DPP4-I are compared with each other or with placebo. Our network meta-analysis presents the associations of DPP4-I versus placebos on HbA1c, FPG, 2 h postprandial blood glucose (2HPPG), BMI, body weight and adverse events. DPP4-I have a lowering effect on the glycemic level (HbA1c, FPG), especially vildaglipti50BID and linagliptin10QD, respectively. Besides, linagliptin5QD has the greatest probabilities of reducing BMI. In addition, DPP4-I were associated with not increasing the incidence of adverse events. Among them, vildagliptin100QD and sitagliptin100QD have the lowest probability in reducing the incidence of hypoglycemia and URTI, respectively.