The SOUL trial (Semaglutide Oral Cardiovascular Outcomes in People with Type 2 Diabetes) results have been presented, and they represent a watershed moment for oral GLP-1 receptor agonist therapy. Let me walk through the design and key findings.[1]
Trial design:
- Multinational, randomized, double-blind, placebo-controlled CVOT
- N = 9,650 patients with T2DM and established atherosclerotic CVD (83%) or CKD (17%)
- Oral semaglutide 14 mg once daily vs placebo, on top of standard of care
- Median follow-up: ~49 months (longest follow-up of any GLP-1 RA CVOT)
Primary endpoint — 3-point MACE:
- HR 0.86 (95% CI: 0.77-0.96), p = 0.006
- Absolute risk reduction: 1.4 percentage points
- NNT: ~71 over 4 years
Individual components:
- CV death: HR 0.93 (95% CI: 0.80-1.09)
- Non-fatal MI: HR 0.74 (95% CI: 0.61-0.89)
- Non-fatal stroke: HR 0.88 (95% CI: 0.70-1.11)
Key secondary:
- All-cause mortality: HR 0.89 (95% CI: 0.78-1.01) — borderline
- Composite kidney outcome: HR 0.78 (95% CI: 0.63-0.97) — significant renal benefit
So oral semaglutide 14mg now has a CVOT demonstrating superiority — joining injectable liraglutide (LEADER) and injectable semaglutide (SUSTAIN-6, SELECT) in the pantheon of GLP-1 RAs with proven cardiovascular benefit. This matters enormously for the oral formulation.
Two discussion points: (1) How does this compare to the injectable sema CVOT data? (2) What does this mean for oral GLP-1 RA positioning in clinical practice?
[1] Husain M, et al. SOUL Trial: Oral Semaglutide and Cardiovascular Outcomes. N Engl J Med. 2025.
Let me address the comparison question directly. There's a temptation to compare SOUL (oral sema 14mg, HR 0.86) unfavorably against SELECT (SC sema 2.4mg, HR 0.80). But that comparison is deeply flawed for several reasons:
- Different populations: SOUL enrolled T2DM patients; SELECT enrolled non-diabetic obesity patients with established CVD. Diabetes itself modifies the CV risk profile and response to interventions.
- Different doses: Oral sema 14mg achieves plasma semaglutide exposures roughly equivalent to SC sema 0.5-1.0mg, NOT 2.4mg. This is a pharmacokinetics issue — oral bioavailability of semaglutide with SNAC is only ~1%. The 14mg oral dose was the maximum approved for T2DM when SOUL was designed.
- Different background therapy: SOUL patients were on more intensive glucose-lowering therapy at baseline, including some on SGLT2 inhibitors (which themselves have CV benefit). Incremental benefit on top of SGLT2i is harder to demonstrate.
- Different follow-up: SOUL had longer follow-up (~4 years vs ~3.3 years), which matters for time-dependent hazard ratios.
The more relevant comparison is SOUL vs LEADER (liraglutide CVOT): LEADER showed HR 0.87 for 3P-MACE, almost identical to SOUL's 0.86. Given that liraglutide and oral semaglutide 14mg achieve similar GLP-1R engagement levels, this consistency is pharmacologically reassuring.[2]
[2] Marso SP, et al. LEADER: Liraglutide and Cardiovascular Outcomes. N Engl J Med. 2016;375(2):311-322.
The oral bioavailability issue deserves deeper exploration because it has major implications for the future.
Current oral semaglutide (Rybelsus) uses the SNAC absorption enhancer (sodium N-[8-(2-hydroxybenzoyl) amino] caprylate), which locally and transiently raises gastric pH and promotes transcellular absorption in the stomach. Even with SNAC, absolute bioavailability is approximately 0.4-1.0%, with enormous inter- and intra-individual variability (CV% for Cmax ~80-100%).[3]
This means that oral sema 14mg gives exposure equivalent to roughly SC sema 0.5-1.0mg. To achieve exposure equivalent to SC sema 2.4mg (the obesity dose), you'd theoretically need oral doses of ~50-70mg, which raises concerns about SNAC-related GI effects and cost.
Novo has been developing higher oral sema doses (25mg and 50mg) with improved formulations. Phase 3 data for oral sema 50mg in obesity (OASIS trials) showed ~17% weight loss at 68 weeks — getting closer to injectable performance but still somewhat lower.[4]
SOUL used the 14mg dose because that's what was approved when the trial was designed. The higher oral doses weren't available. So in some sense, SOUL demonstrates CV benefit with a "subtherapeutic" dose relative to what we now know is achievable orally.
[3] Buckley ST, et al. Transcellular stomach absorption of semaglutide. Sci Transl Med. 2018;10(467):eaar7047.
[4] Knop FK, et al. Oral semaglutide 50mg in obesity (OASIS 1). Lancet. 2023;402:705-719.
PeptideDetective — Independent Peptide Analytics
Community-driven peptide testing and vendor rating platform. Transparent results. Unbiased analysis. Trusted by thousands.
View ResultsFrom a clinical practice standpoint, SOUL is hugely important for one simple reason: many patients refuse injections. Full stop. I have patients who are appropriate candidates for GLP-1 RAs but will not consider a weekly subcutaneous injection under any circumstances — needle phobia, convenience concerns, perceived stigma.
Before SOUL, I could offer these patients oral semaglutide for glycemic control but couldn't cite a CVOT for cardiovascular benefit with the oral formulation. Now I can. This changes the conversation.
Practical challenges with oral sema that SOUL doesn't address:
- Must be taken fasting with <=4 oz water, 30 min before eating — adherence to dosing instructions is poor in real-world settings
- Drug interactions with PPIs (reduce absorption) — many T2DM/CVD patients are on PPIs
- Highly variable absorption means some patients get excellent exposure and others get minimal
- The 14mg dose provides less weight loss than SC 2.4mg, which matters for patients with dual obesity/CVD indications
I'm hoping the next-gen oral formulations (higher doses, improved absorption technology) will solve some of these issues. But for now, having CVOT data for the oral formulation is a significant clinical win.
Can someone explain what a CVOT actually is and why it matters so much? I keep seeing the term but I'm not clear on why it's such a big deal for a diabetes drug to have one.