Glucometabolic outcomes of GLP-1 receptor agonist-based therapies in patients with type 2 diabetes: a systematic review and network meta-analysis

Abstract

Background: Innovative GLP-1 receptor agonist (GLP-1RA)-based treatment strategies-such as tirzepatide, GLP-1RA plus basal insulin fixed-ratio combinations [FRC], GLP-1RA plus sodium glucose cotransporter-2 inhibitors [SGLT-2i] combinations, and high-dose GLP-1RA-have been listed among the most efficacious options for type 2 diabetes management. However, differences in their glucometabolic effects have not been assessed in dedicated head-to-head trials. In the absence of such trials, we aimed to provide a useful comparison among these treatment strategies to guide clinical practice.

Methods: In this network meta-analysis, we searched PubMed, MEDLINE, and Web of Science (from database inception to June 24, 2023) for randomised controlled studies, published in English, that enrolled individuals with type 2 diabetes treated with tirzepatide, iGlarLixi, iDegLira, GLP-1RA plus SGLT-2i combination, or high-dose GLP-1RA (dulaglutide 3 mg and 4.5 mg, semaglutide 2 mg) compared with placebo or active comparators. Eligible studies reported change from baseline in HbA1c as an outcome, which was the primary outcome of this analysis. Secondary outcomes were changes in fasting and post-prandial glucose, bodyweight, LDL-cholesterol, blood pressure and risk of hypoglycaemia. We assessed risk of bias through the Cochrane Collaboration's tool (RoB2 tool), publication bias through visual inspection of funnel plots and Egger's test, and heterogeneity by comparing the magnitude of the common between-study variance (τ²) for each outcome with empirical distributions of heterogeneity variances. This network meta-analysis was registered in PROSPERO (CRD42022329878).

Findings: 40 trials were included. Tirzepatide 15 mg ranked first in terms of HbA1c reduction compared to other GLP-1RA-based strategies, even those including insulin (vs. iDegLira MD -0.40%, 95% CI [-0.66; -0.14], low certainty; vs. iGlarLixi MD -0.48%, 95% CI [-0.75; -0.21], low certainty), without increasing the risk of hypoglycaemia (vs. iDegLira OR 0.35, 95% CI [0.16; 0.79], high certainty; vs. iGlarLixi OR 0.31, 95% CI [0.20; 0.48], high certainty). Tirzepatide 15 mg was also the most efficacious on weight lowering, even compared to high-dose GLP-1RA (eg, semaglutide 2 mg MD -6.56 kg, 95% CI [-7.38; -5.73], low certainty) and GLP-1RA plus SGLT-2i combination (MD -4.61 kg, 95% CI [-5.29; -3.93], low certainty). Risk of bias and publication bias were generally low throughout studies, while high levels of heterogeneity were detected for most outcomes.

Interpretation: Aiming to support clinicians in tailoring treatment to patients' needs, we suggest that a hierarchy among treatment strategies be devised considering the best options for type 2 diabetes. Tirzepatide, followed by GLP-1RA plus basal insulin FRC and GLP-1RA plus SGLT-2i combination, was associated with greater benefit on HbA1c than high-dose GLP-1RA.

Funding: Fondazione per la Ricerca Biomedica "Saverio e Isabella Cianciola" and Next Generation EU, in the context of the National Recovery and Resilience Plan, Investment PE8-Project Age-It: Ageing Well in an Ageing Society.

Keywords: Fixed-ratio combination; GLP1-RA; Network meta-analysis; SGLT-2i; Tirzepatide.

Fig. 1

PRISMA flowchart for study selection.

Fig. 2

Meta-analysis networks for change in HbA1c level. Each circle indicates a treatment node, and its size is proportional to the number of trials evaluating each treatment. Lines connecting two nodes represent direct comparisons between two treatments; the thickness of the lines is proportional to the number of trials directly comparing the 2 connected treatments. BIAsp30, biphasic insulin aspart 30/70; GLP-1RA, glucagon-like peptide-1 receptor agonists; SGLT-2i, sodium glucose cotransporter-2 inhibitors; SU, sulfonylurea.

Fig. 3

Network meta-analysis results for change from baseline in a. HbA1c, b. fasting plasma glucose (FPG), and c. post-prandial glucose (PPG) compared with placebo. Treatments are presented according to their effect estimate compared with placebo. Effect sizes are presented as mean difference (MD) and 95% confidence intervals (CI). New GLP-1RA-based treatments are highlighted in black, other treatments in grey. SGLT-2i and GLP-1RA compounds and doses in included studies are described in Supplementary Appendix S2. BIAsp30, biphasic insulin aspart 30/70; GLP-1 RA, glucagon-like peptide-1 receptor agonists; SGLT-2i, sodium glucose cotransporter-2 inhibitors; SU, sulfonylurea.

Fig. 4

Network meta-analysis results for change from baseline in bodyweight (BW) compared with placebo. Treatments are presented according to their effect estimate compared with placebo. Effect sizes are presented as mean difference (MD) and 95% confidence intervals (CI). New GLP-1RA-based treatments are highlighted in black, other treatments in grey. BIAsp30, biphasic insulin aspart 30/70; GLP-1 RA, glucagon-like peptide-1 receptor agonists; SGLT-2i, sodium glucose cotransporter-2 inhibitors; SU, sulfonylurea.

Fig. 5

Network meta-analysis results for hypoglycaemia compared with placebo. Treatments are presented according to their effect estimate compared with placebo. Effect sizes are presented as odds ratio (OR) and 95% confidence intervals (CI). New GLP-1RA-based treatments are highlighted in black, other treatments in grey. BIAsp30, biphasic insulin aspart 30/70; GLP-1 RA, glucagon-like peptide-1 receptor agonists; SGLT-2i, sodium glucose cotransporter-2 inhibitors; SU, sulfonylurea.