Network Meta-Analysis: Distinct Benefit–Risk Profiles of GLP-1 Receptor Agonists, DPP-4 Inhibitors, and SGLT2 Inhibitors

Background: Type 2 diabetes (T2D) is a global health challenge due to its high prevalence and associated risks for cardiovascular (CV) and renal complications. Newer glucose-lowering drug (GLD) classes—dipeptidyl peptidase 4 inhibitors (DPP4i), glucagon-like peptide-1 receptor agonists (GLP-1RAs), and sodium-glucose cotransporter 2 inhibitors (SGLT2i)—offer unique benefits and safety profiles. Although many large randomized outcome trials have demonstrated their efficacy, the benefit–risk balance among these agents remains incompletely understood, especially regarding non-traditional outcomes such as psychiatric disorders and neurodegenerative diseases. Given their generally higher costs and frequent industry sponsorship of the trials, it is critical to evaluate these classes in a systematic and impartial manner.

Objective: To systematically compare the benefits and risks of DPP4i, GLP-1RAs, and SGLT2i in adults with T2D, heart failure, or chronic kidney disease, focusing on 21 outcomes spanning macrovascular and microvascular events, infections, psychiatric outcomes, and cancer risk.

Methods: Researchers searched PubMed, Embase, and CENTRAL through November 2023 for randomized placebo-controlled cardiovascular and kidney outcome trials of DPP4i, GLP-1RAs, or SGLT2i in adults with T2D, heart failure, or chronic kidney disease. Twenty-six trials (N=198,177 participants) met inclusion criteria. A novel PAtient-centered treatment ranking via Large-scale Multivariate network meta-analysis (PALM) approach was used to synthesize population-averaged odds ratios (ORs) with 95% confidence intervals (CIs). This approach allowed for the simultaneous evaluation of multiple outcomes and the generation of weighted origami plots to visualize treatment rankings across different disease categories. Heterogeneity (I²) and publication bias (Egger’s test) were also assessed.

Results:

• Macrovascular: GLP-1RAs reduced major adverse cardiovascular events (MACE) versus placebo (OR 0.85, 95% CI 0.79–0.92) and DPP4i. GLP-1RAs also lowered stroke risk (OR range 0.82–0.85 vs comparators). SGLT2i yielded the largest reduction in hospitalization for heart failure (OR 0.68, 95% CI 0.64–0.73 vs placebo). DPP4i had a lower amputation risk relative to SGLT2i (OR 0.85, 95% CI 0.75–0.95).
• Microvascular: SGLT2i most effectively reduced renal composite outcomes versus DPP4i and placebo (OR ~0.67). However, DPP4i was linked to higher neuropathy risk (OR 1.10, 95% CI 1.02–1.18 vs placebo).
• Psychiatric and Neurodegenerative: DPP4i was associated with a reduced risk of depression, suicide, and alcohol use disorder compared to placebo. A lower Parkinson’s disease risk (OR 0.54, 95% CI 0.32–0.92) was also noted. GLP-1RAs showed a possible increased risk of suicidal ideation vs DPP4i, though evidence remains inconclusive.
• Infections/Inflammation: SGLT2i substantially increased genital infections (OR 3.11, 95% CI 2.15–4.50 vs placebo). DPP4i had higher pancreatitis risk vs all comparators.
• Cancer: GLP-1RAs were linked to increased thyroid cancer risk (OR range 1.58–2.70), though overall cancer risk and pancreatic cancer did not differ significantly across treatments.

Conclusions: Each newer GLD class offers specific advantages along with distinct safety considerations. GLP-1RAs appear particularly effective for macrovascular endpoints but carry a signal for thyroid malignancy. SGLT2i provide notable cardioprotective and renoprotective effects, offset by risk of genital infections and an increased risk of amputation compared to DPP4i. DPP4i tend to have neutral CV/renal outcomes yet may protect against certain psychiatric issues and neurodegenerative conditions, balanced by increased pancreatitis risk. Personalized treatment strategies, with attention to comorbidities and adverse event profiles, remain essential—particularly given the generally elevated costs of these newer agents. It is important to note that these findings are largely based on indirect comparisons in the absence of head-to-head trials, which is a limitation.

Implications for Practice: Clinicians should weigh cardiovascular risk, kidney function, mental health status, and potential malignancy when prescribing GLDs. SGLT2i may be favored in patients with heart failure or chronic kidney disease, while GLP-1RAs may be ideal for those with high atherosclerotic CV risk. DPP4i could be considered in patients with psychiatric or neurologic comorbidities but require caution regarding pancreatitis. Cost considerations and access may also influence real-world use.

Study Strengths and Limitations: Strengths include a large number of participants, broad outcome coverage, and a multivariate network meta-analysis that accommodates indirect comparisons. Limitations arise from underreported outcomes (e.g., psychiatric, neurodegenerative), heterogeneous trial populations, potential publication bias for some outcomes, and possible off-target influences not fully captured. Furthermore, sponsor involvement in all included trials warrants cautious interpretation of benefit–risk claims.

Future Research: Head-to-head trials comparing newer GLDs for psychiatric and neurologic endpoints, along with detailed reporting of rare adverse events (e.g., pancreatitis, cancer subtypes), are needed. Studies on real-world cost-effectiveness and access issues could clarify how to optimize therapy in routine practice. Additional investigations into long-term safety signals (including suicidality and thyroid malignancies) would further guide clinical decision-making.

Reference: Tang H, Zhang B, Lu Y, Donahoo WT, Singh Ospina N, Kotecha P, Lu Y, Tong J, Smith SM, et al. “Assessing the benefit–risk profile of newer glucose-lowering drugs: A systematic review and network meta-analysis of randomized outcome trials.” Diabetes, Obesity and Metabolism. First published: 26 December 2024. DOI: http://doi.org/10.1111/dom.16147