Evaluation and comparison of efficacy and safety of tirzepatide and semaglutide in patients with type 2 diabetes mellitus: A Bayesian network meta-analysis

As new antidiabetic drugs, tirzepatide (Tir) and semaglutide (Sem) are progressively applied in clinical practice. However, their efficacy and safety profiles have not been comprehensively assessed. Therefore, a Bayesian network meta-analysis was used to evaluate and compare the efficacy and safety of Tir and Sem in treating type 2 diabetes mellitus (T2DM). PubMed, EMBASE, Web of Science, Cochrane Library and ClinicalTrials.gov were systematically searched from inception to April 3rd, 2023. Randomized clinical trials (RCTs) comparing the efficacy and safety of Tir and Sem with placebo or the other antidiabetic drugs in treating T2DM were included. The efficacy outcomes included changes in glycated hemoglobin (HbA1c), body weight (BW), body mass index (BMI), and the proportion of participants with HbA1c<7%. The safety outcome was the proportion of participants experiencing gastrointestinal adverse events (GIAEs). A total of 38 studies involving 34,166 participants were included. Compared to 1 mg of subcutaneous Sem (Sem SC), 5 mg, 10 mg and 15 mg of Tir demonstrated superior efficacy in reducing HbA1c (mean difference (MD), [95% CI], -0.22 [-0.40, -0.03] %, -0.42 [-0.60, -0.24] % and -0.53 [-0.71, -0.35] %, respectively) and BW (MD [95% CI], -1.48 [-2.53, -0.43] kg, -4.00 [-5.05, -2.95] kg and -5.71 [-6.73, -4.68] kg, respectively). Conversely, 7 mg and 14 mg of oral Sem (Sem PO) displayed inferior efficacy in reducing HbA1c (MD [95% CI], 0.51 [0.29, 0.73] % and 0.39 [0.20, 0.57] %, respectively) and BW (MD [95% CI], 2.36 [1.24, 3.48] kg and 1.11 [0.10, 2.13] kg). However, 20 mg and 40 mg of Sem PO were non-inferior in reducing HbA1c (MD [95% CI], 0.13 [-0.29, 0.55] % and 0.01 [-0.38, 0.40] %, respectively) and BW (MD [95% CI], -0.41 [-2.71, 1.90] kg and -1.32 [-3.58, 0.92] kg). In terms of safety, compared to 1 mg of Sem SC, 5 mg, 10 mg and 15 mg of Tir did not significantly increase the incidence of GIAEs (odd ratio (OR) [95% CI], -0.59 [0.30, 1.1], 0.88 [0.45, 1.7] and 1.1 [0.59, 2.1], respectively), while 7 mg of Sem PO showed a lower incidence of GIAEs (OR [95% CI], 0.35 [0.16, 0.78]). Compared to insulin, 0.5 mg of Sem SC, 1 mg of Sem SC, 5 mg of Tir, 10 mg of Tir and 15 mg of Tir displayed better efficacy in lowering HbA1c (MD [95% CI], -0.40 [-0.63, -0.18] %, -0.69 [-0.90, -0.48] %, -0.91 [-1.10, -0.72] %, -1.11 [-1.30, -0.92] % and 1.22 [-1.41, -1.03] %, respectively) and BW (MD [95% CI], -5.33 [-6.59, -4.1] kg, -6.69 [-7.89,-5.51] kg, -8.17 [-9.26, -7.1] kg, -10.7 [-11.78, -9.62] kg and -12.42 [-13.49,-11.34] kg, respectively). According to the surface under the cumulative ranking curve (SUCRA) value, among all the included interventions, 15 mg of Tir exhibited the most potent effect in reducing HbA1c (99.81%) and BW (99.98%), followed by 10 mg of Tir (96.83% and 95.72%), 5 mg of Tir (92.88% and 86.04%), 1 mg of Sem SC (85.85% and 74.97%), 40 mg of Sem PO (83.36% and 84.31%), 20 mg of Sem PO (76.98% and 77.12%), 300 mg of Can (49.93% and 60.89%), insulin (36.38% and 0.22%) and 100 mg of Sit (12.28% and 18.51%) respectively. Meanwhile, 5 mg, 10 mg, and 15 mg of Tir (48.32%, 30.96%, and 21.07%, respectively), 0.5 mg and 1 mg of Sem SC (33.54% and 24.77%, respectively) significantly increased the incidence of GIAEs. Both Tir and Sem demonstrated favorable antidiabetic effects and were particularly suitable for T2DM patients who were obese or overweight. Despite a high incidence of GIAEs, their safety profile was deemed acceptable. Tir was the best option among all the included interventions.