Cross-trial comparisons are the junk food of evidence-based medicine — they look satisfying but don't nourish. Different patient populations, different study designs, different endpoints, different titration schedules all confound indirect comparisons. SURMOUNT-5 eliminates these confounders by putting both drugs in the same trial. What does a clean comparison actually show?
Study design
SURMOUNT-5 was an open-label, randomized Phase 3b trial comparing tirzepatide 15mg vs semaglutide 2.4mg in adults with obesity (BMI ≥30) or overweight (BMI ≥27 with comorbidities). The primary endpoint was percent change in body weight at 72 weeks. Secondary endpoints included the proportion achieving ≥5%, ≥10%, ≥15%, and ≥20% weight loss, plus waist circumference and cardiometabolic parameters.
Open-label design means both patients and investigators knew which drug was being administered. This introduces potential bias — particularly for subjective endpoints and adherence behavior — though it's standard for active-comparator trials in this class.
The results: tirzepatide wins on weight
Tirzepatide 15mg produced approximately 6 percentage points more weight loss than semaglutide 2.4mg. This difference was statistically significant and clinically meaningful. The proportion achieving ≥20% weight loss was substantially higher with tirzepatide — roughly 45-50% vs 20-25%.
| Endpoint | Tirzepatide 15mg | Semaglutide 2.4mg |
|---|---|---|
| Mean weight loss (%) | ~20% | ~14% |
| ≥5% weight loss | ~93% | ~85% |
| ≥10% weight loss | ~82% | ~63% |
| ≥15% weight loss | ~65% | ~40% |
| ≥20% weight loss | ~47% | ~22% |
What SURMOUNT-5 doesn't prove
It doesn't prove tirzepatide is "better" for every patient. The comparison is max-dose tirzepatide (15mg) against standard-dose semaglutide (2.4mg). Semaglutide 7.2mg — now in regulatory review — narrows the gap significantly. A SURMOUNT-5 sequel comparing tirzepatide 15mg vs semaglutide 7.2mg would be a different story.
It doesn't account for tolerability in real-world practice. The trial enrolled participants who tolerated the titration to full dose. Patients who discontinued due to GI intolerance are not represented in the efficacy results. Real-world completion rates may differ between the two drugs.
It doesn't address long-term safety differences. 72 weeks is sufficient to measure weight loss. It is not sufficient to detect differences in rare long-term safety signals (pancreatitis, thyroid risk, cardiovascular events). SELECT and SURPASS-CVOT address long-term safety for semaglutide and tirzepatide separately, but no long-term head-to-head safety comparison exists.
It doesn't mean semaglutide is inadequate. Semaglutide 2.4mg produced 14% mean weight loss — a clinically significant result by any measure. The comparison to tirzepatide's 20% shows that tirzepatide is more effective, not that semaglutide is ineffective. For patients who achieve their treatment goals on semaglutide, switching to tirzepatide for incremental improvement may not be warranted.
SURMOUNT-5 confirms what cross-trial estimates suggested: tirzepatide 15mg produces more weight loss than semaglutide 2.4mg. The ~6 percentage point difference is real and statistically significant. But the clinical decision between drugs depends on individual response, tolerability, insurance coverage, cost, and whether the patient's goals are met by semaglutide alone. SURMOUNT-5 informs the decision — it doesn't make it.