PDF: SOUL
(2025)Objective
Semaglutide once-daily oral pills vs placebo in patients with type 2 diabetes and atherosclerotic cardiovascular disease, chronic kidney disease, or both.
Study Summary
• Oral semaglutide significantly reduced major adverse cardiovascular events (MACE) compared to placebo (HR 0.86; P=0.006), mainly driven by a reduction in nonfatal myocardial infarction with favorable safety profile.
Intervention
Oral semaglutide (max 14 mg/day) vs. placebo, in addition to standard of care, among patients with type 2 diabetes and ASCVD and/or CKD.
Inclusion Criteria
Age ≥50, type 2 diabetes (HbA1c 6.5–10.0%), and established ASCVD, CKD (eGFR <60), or both. Excluded if ESRD or on chronic kidney replacement therapy.
Study Design
Arms: Oral Semaglutide vs. Placebo
Patients per Arm: Semaglutide: 4825, Placebo: 4825
Outcome
• Primary outcome (MACE: CV death, MI, stroke): 12.0% vs. 13.8% (HR 0.86; P=0.006). <br>• Nonfatal MI: 4.0% vs. 5.2% (HR 0.74). <br>• Nonfatal stroke: 3.0% vs. 3.3% (HR 0.88). <br>• CV death: 6.2% vs. 6.6% (HR 0.93). <br>• Major kidney disease events: 8.4% vs. 9.0% (HR 0.91; P=0.19). <br>• Major limb events: 1.5% vs. 2.1% (HR 0.71). <br>• All-cause death: 10.9% vs. 12.0% (HR 0.91). <br>• Serious adverse events: 47.9% (semaglutide) vs. 50.3% (placebo). <br>• GI discontinuation: 6.4% vs. 2.0%.
Bottom Line
Among high-risk patients with type 2 diabetes, once-daily oral semaglutide was superior to placebo in reducing the risk of major adverse cardiovascular events over a median follow-up of 49.5 months. Serious adverse events were also lower in the semaglutide group.
Major Points
- Largest and longest GLP-1 RA cardiovascular outcomes trial: 9,650 patients with T2DM and established ASCVD, CKD, or both, followed for median 49.5 months across 444 centers in 33 countries.
- Oral semaglutide reduced 3-point MACE by 14% vs placebo (12.0% vs 13.8%, HR 0.86, 95% CI 0.77–0.96, P=0.006) — first oral GLP-1 RA to demonstrate cardiovascular superiority.
- Builds on injectable semaglutide results (SUSTAIN 6, SELECT) — confirms the GLP-1 RA class effect extends to oral formulation, expanding access for injection-averse patients.
- NNT of 50 over 3 years (95% CI 31–125) to prevent one MACE event — clinically meaningful absolute risk reduction of 1.8%.
- First confirmatory secondary endpoint (major kidney disease events) did not reach significance (HR 0.91, P=0.19), halting the hierarchical testing procedure — kidney benefit remains unproven.
- CV death not significantly reduced (HR 0.93, P not tested due to hierarchical failure) — the MACE benefit was driven by nonfatal MI and nonfatal stroke reduction rather than mortality.
- All-cause mortality showed a favorable trend (HR 0.91, 95% CI 0.80–1.02) but was not formally tested — tantalizing but not conclusive for a mortality benefit.
- Higher discontinuation in semaglutide group (15.5% vs 11.6%) due to GI side effects — raises concern about real-world adherence and durability of benefit.
- Substantial SGLT2 inhibitor co-use at baseline (26.9%) — results demonstrate additive benefit on top of modern diabetes therapy including SGLT2i.
- Underrepresentation of women (28.9%) and Black patients (2.6%) limits generalizability — the trial population was predominantly white and male with established CVD.
Study Design
- Study Type
- International, double-blind, randomized, placebo-controlled, event-driven, superiority phase 3b trial.
- Randomization
- Yes
- Blinding
- Double-blind.
- Sample Size
- 9650
- Follow-up
- Median of 49.5 months.
- Centers
- 444
- Countries
- 33 countries
Primary Outcome
Definition: Major adverse cardiovascular events (a three-point composite of death from cardiovascular causes, nonfatal myocardial infarction, or nonfatal stroke).
| Control | Intervention | HR/OR | P-value |
|---|---|---|---|
| 13.8% (668/4825) | 12.0% (579/4825) | 0.86 (0.77 to 0.96) | 0.006 |
Limitations & Criticisms
- Limited generalizability — enrolled only patients with established ASCVD, CKD, or both; results do not apply to primary prevention or lower-risk T2DM populations.
- Severe underrepresentation of women (28.9%) and Black patients (2.6%) — cardiovascular outcomes may differ in these populations.
- Failure of the kidney endpoint (HR 0.91, P=0.19) halted hierarchical testing — CV death and all-cause mortality benefits cannot be formally claimed despite favorable trends.
- Higher discontinuation in semaglutide group (15.5% vs 11.6%) primarily due to GI side effects — real-world adherence may be lower than trial setting, attenuating benefit.
- Industry-sponsored by Novo Nordisk (semaglutide manufacturer) — potential bias in trial design, conduct, and interpretation.
- 14% relative risk reduction (HR 0.86) is modest compared to other CV interventions — clinical significance debated given the cost of oral semaglutide and the NNT of 50.
- Cannot distinguish whether the CV benefit is mediated by glucose lowering, weight loss, anti-inflammatory effects, or direct vascular mechanisms — the mechanism remains unclear.
- Substantial background therapy including SGLT2i (26.9%) and statins — the incremental benefit of adding semaglutide to optimized modern therapy is relatively small.
- Oral formulation requires fasting administration (30 minutes before first food/drink) — practical adherence challenges not captured in the trial's supervised setting.
Citation
N Engl J Med 2025