Why didn't SURPASS-2 include semaglutide 2 mg as a comparator?

Eli Lilly stated that semaglutide 2 mg was not yet approved when the protocol was locked. While technically correct, the 2 mg dose had phase 3 data available during the design phase. Its absence means the trial compared maximally titrated tirzepatide against a mid-range semaglutide dose, limiting the clinical relevance of the superiority claim for the lowest tirzepatide arm.

Was SURPASS-2 blinded?

No. SURPASS-2 was an open-label trial. Participants and investigators knew which drug was administered. This was necessary because tirzepatide and semaglutide use different injection devices. The primary endpoint (A1C) is an objective lab measurement, which reduces the impact of open-label bias on the main results.

How diverse was the SURPASS-2 study population?

Approximately 73% of participants were white. Black, Hispanic, and Asian patients were underrepresented relative to the global burden of type 2 diabetes. This limits confidence in applying the results to populations with different genetic, metabolic, and socioeconomic profiles.

Does SURPASS-2 prove tirzepatide is better for cardiovascular outcomes?

No. SURPASS-2 was a glycemic efficacy trial, not a cardiovascular outcomes trial. Semaglutide has proven cardiovascular benefit from SUSTAIN-6. Tirzepatide's cardiovascular outcomes data (SURPASS-CVOT) are still maturing. Clinicians should not extrapolate A1C superiority to cardiovascular superiority.

What was the dropout rate in SURPASS-2?

Overall discontinuation was roughly 10 to 15% across arms, with 3 to 7% discontinuing due to adverse events. These rates are within the expected range for injectable GLP-1 receptor agonist trials, though the open-label design may have influenced adherence differently than a blinded study would.

Were patients on insulin or SGLT2 inhibitors included?

No. SURPASS-2 enrolled patients on metformin monotherapy only. This cleaner background regimen likely amplified the observed A1C reductions compared to what would be seen in patients already taking multiple glucose-lowering agents, which is the more common clinical scenario.

Who funded SURPASS-2?

Eli Lilly funded the trial, designed the protocol, conducted the statistical analysis, and employed several of the listed authors. Standard conflict-of-interest disclosures were published. Independent replication or investigator-initiated head-to-head studies would strengthen the evidence base.

Is 40 weeks long enough to judge these drugs?

Forty weeks is sufficient to capture peak glycemic and weight effects but does not address long-term durability, weight regain patterns, or rare adverse events. Extension studies and real-world evidence are needed to fill that gap.

How should clinicians interpret the A1C superiority margin?

The margin of superiority for tirzepatide 5 mg over semaglutide 1 mg was 0.15 percentage points. While statistically significant in a well-powered trial, this is a clinically modest difference. The 10 mg and 15 mg doses showed larger margins (0.28 and 0.44 points), which carry more clinical weight.

Did any independent commentary challenge the trial's conclusions?

Yes. Letters to the NEJM and subsequent review articles questioned the comparator dose selection, the lack of racial diversity, and the absence of cardiovascular or cost-effectiveness endpoints. The ADA's 2023 Standards of Care acknowledged SURPASS-2 while noting the need for data against maximally dosed semaglutide.

Honest Criticisms and Limitations of the SURPASS-2 Trial

By HealthRX.com Medical Team

Published May 25, 2026Updated May 25, 2026Last reviewed May 25, 2026

Clinical image for Honest Criticisms and Limitations of the SURPASS-2 Trial Image: HealthRX.com clinical image

At a glance

| Detail | Value | |---|---| | Trial | SURPASS-2 (NCT04158830) | | N | 1,879 | | Intervention | Tirzepatide 5 mg, 10 mg, or 15 mg weekly | | Comparator | Semaglutide 1 mg weekly | | Duration | 40 weeks | | Primary endpoint | Change in A1C from baseline | | Key result | All three tirzepatide doses were superior to semaglutide 1 mg for A1C reduction (−2.01% to −2.30% vs −1.86%) and body weight loss |

The Comparator Problem: Why 1 mg Semaglutide?

The single most debated design choice in SURPASS-2 is the semaglutide comparator dose. Semaglutide (Ozempic) is approved at 0.5 mg, 1 mg, and 2 mg for type 2 diabetes. The 2 mg dose, which the FDA approved in March 2022, was already in late-stage development when SURPASS-2 was designed. By selecting only the 1 mg dose, the trial compared three escalating tirzepatide doses (5, 10 to 15 mg) against a mid-range semaglutide dose rather than the highest available option.

This matters arithmetically. In the SUSTAIN trials, semaglutide 2 mg produced roughly 0.2 percentage points more A1C reduction than the 1 mg dose. The margin of superiority for tirzepatide 5 mg over semaglutide 1 mg in SURPASS-2 was 0.15 percentage points. A 2 mg semaglutide arm could have narrowed or erased that gap at the lowest tirzepatide dose.

Eli Lilly has stated that semaglutide 1 mg was the highest approved dose at the time the protocol was finalized. That is technically accurate. But clinical trialists routinely include doses under regulatory review when the timeline permits, and the 2 mg dose had phase 3 data available by late 2019. The absence of a 2 mg arm does not invalidate the trial, but it makes the "superiority" framing incomplete for clinicians choosing between a maximally titrated GLP-1 and tirzepatide.

Enrollment and Baseline Characteristics

SURPASS-2 enrolled adults with type 2 diabetes on metformin monotherapy, with baseline A1C between 7.0% and 10.5%. Mean baseline A1C was approximately 8.3%, and mean BMI was around 34 kg/m². The population was relatively young (mean age ~57), predominantly white (~73%), and had a mean diabetes duration of about 8.6 years.

Several enrollment filters limit generalizability:

| Exclusion | Clinical consequence | |---|---| | eGFR <45 mL/min/1.73 m² | Excludes moderate-to-severe CKD, common in long-standing T2D | | Prior bariatric surgery | Removes patients with surgically altered GI physiology | | NYHA Class III, IV heart failure | Excludes advanced HF, a frequent T2D comorbidity | | History of pancreatitis | Standard for GLP-1 trials but limits real-world applicability | | Insulin or SGLT2i use | Patients were on metformin only, a less complex regimen than most real-world T2D patients |

The metformin-only requirement is worth highlighting. In routine practice, many patients considered for a GLP-1 receptor agonist or dual GIP/GLP-1 agonist are already on two or three oral agents, sometimes including an SGLT2 inhibitor with proven cardiovascular benefit. SURPASS-2's cleaner background regimen may have amplified the observed A1C effect for both arms compared to what clinicians see in more complex patients.

Duration: 40 Weeks Is Not Long Enough

Forty weeks captures the glycemic nadir for most incretin-based therapies, but it tells us little about durability. Weight regain after GLP-1 discontinuation is well documented. The STEP 1 extension data showed that participants regained roughly two-thirds of lost weight within one year of stopping semaglutide 2.4 mg.

SURPASS-2 was not designed to answer whether tirzepatide's weight advantage persists at 18 or 24 months, or whether the dual incretin mechanism produces a different durability curve than GLP-1 alone. The 40-week snapshot favors whichever agent reaches its peak effect faster. Longer-duration data from SURMOUNT-1 (72 weeks with tirzepatide for obesity) and SURPASS-4 (up to 104 weeks in a cardiovascular safety context) partially address this gap, but neither used semaglutide as a direct comparator.

Statistical Design: Multiplicity and the Estimand

The trial used a modified intention-to-treat analysis with a treatment-policy estimand for the primary endpoint. This approach includes all randomized patients regardless of adherence. The authors also reported an efficacy estimand that excluded data after treatment discontinuation or rescue medication use.

Between these two approaches, the results shift. Under the efficacy estimand, tirzepatide 15 mg reduced A1C by 2.37% versus 1.86% for semaglutide. Under the treatment-policy estimand, the gap narrowed to 2.30% versus 1.86%. Both showed statistical superiority, but the different framing can influence how results are presented in promotional materials versus clinical guidelines.

Multiplicity adjustment is another detail worth noting. The trial tested three tirzepatide doses against one comparator across multiple endpoints (A1C, weight, proportion reaching A1C <7%, proportion reaching A1C <5.7%). The prespecified hierarchical testing procedure controlled for family-wise error, but the sheer number of comparisons (twelve primary and key secondary tests) can inflate the perceived strength of evidence if readers focus on individual p-values without understanding the testing cascade.

Gastrointestinal Tolerability: Framing vs. Data

Nausea was the most common adverse event across all tirzepatide arms (12 to 18%) and in the semaglutide arm (18%). The trial reports these rates as comparable. That framing is reasonable for overall incidence, but it obscures dose-dependent patterns within the tirzepatide arms.

| Adverse event | Tirzepatide 5 mg | Tirzepatide 10 mg | Tirzepatide 15 mg | Semaglutide 1 mg | |---|---|---|---|---| | Nausea | 12.2% | 16.4% | 18.3% | 17.9% | | Diarrhea | 11.5% | 13.3% | 11.7% | 8.3% | | Vomiting | 2.8% | 5.3% | 8.3% | 8.3% | | Decreased appetite | 5.5% | 9.1% | 10.0% | 8.7% |

At the 15 mg tirzepatide dose (the one with the largest efficacy margin), GI side effects are roughly equivalent to semaglutide 1 mg. But semaglutide 2 mg typically produces higher GI event rates than 1 mg, so a fairer tolerability comparison would require the higher semaglutide dose in the trial. The absence of that arm means we cannot say whether tirzepatide's tolerability advantage holds when both drugs are maximally dosed.

Discontinuation due to adverse events was low across all arms (3 to 7%), which is reassuring. The study was not powered to detect rare safety signals like acute pancreatitis or medullary thyroid carcinoma risk, events that require much larger populations and longer follow-up.

Conflict of Interest and Funding

SURPASS-2 was designed, funded, and analyzed by Eli Lilly, the manufacturer of tirzepatide (Mounjaro). Multiple authors were Lilly employees. The remaining academic investigators received consulting fees or grants from Lilly and, in some cases, from Novo Nordisk (semaglutide's manufacturer) as well.

This does not mean the data are fabricated. Industry-sponsored trials undergo rigorous regulatory scrutiny, and the raw data supported FDA approval. But sponsor involvement in study design creates structural incentives that are worth naming:

Dose selection favored the sponsor's drug (three doses tested) over the comparator (one dose tested).
Endpoint selection emphasized A1C, where tirzepatide's dual mechanism has a clear pharmacologic advantage, rather than cardiovascular outcomes or patient-reported quality of life.
Publication timing aligned with the Lilly regulatory submission calendar.

The 2021 NEJM publication includes a standard conflict-of-interest disclosure. Readers should weigh these disclosures when interpreting the trial's positioning as definitive evidence of superiority.

What Post-Publication Commentary Raised

Letters to the editor and independent commentaries published after SURPASS-2 highlighted several recurring concerns:

Comparator dose adequacy. Multiple correspondents noted the missing 2 mg semaglutide arm. The American Diabetes Association's 2023 Standards of Care cites SURPASS-2 but notes that head-to-head data against maximally dosed semaglutide are still needed.
Racial and ethnic diversity. With ~73% white enrollment, the trial's applicability to Black, Hispanic, and South Asian populations (groups with disproportionately high T2D burden) remains uncertain. Pharmacogenomic variation in GIP receptor signaling has been hypothesized but not tested in this context.
Cardiovascular outcome gap. SURPASS-2 was a glycemic efficacy trial, not a cardiovascular outcomes trial (CVOT). Semaglutide has completed CVOT data (SUSTAIN-6 showed a significant reduction in major adverse cardiovascular events). Tirzepatide's dedicated CVOT, SURPASS-CVOT, is ongoing. Until those data mature, clinicians treating high-CV-risk patients may reasonably prefer a drug with proven cardiovascular benefit.
Cost-effectiveness silence. The trial reported no pharmacoeconomic data. At U.S. list prices, tirzepatide and semaglutide are both expensive. Payers making formulary decisions need cost-per-A1C-point or cost-per-kg analyses that SURPASS-2 does not provide.

What SURPASS-2 Does Well

Acknowledging limitations should not erase what the trial accomplished. The randomization was sound (1:1:1:1), the sample size was large enough for the primary endpoint, and the 40-week duration is standard for glycemic superiority trials. The open-label design (necessary because of different injection devices) could introduce bias, but A1C is an objective lab value less susceptible to placebo effects than subjective endpoints. The consistent dose-response across all three tirzepatide arms strengthens the biological plausibility of dual GIP/GLP-1 agonism as a mechanism.

The trial answered the question it asked: does tirzepatide at three doses beat semaglutide 1 mg for A1C reduction over 40 weeks in metformin-treated T2D? Yes. The criticism is not that the answer is wrong. The criticism is that the question was crafted in a way that made a favorable answer more likely, and that clinicians need to recognize the gap between the trial's question and their own.

Frequently asked questions

›

References

Frías JP, Davies MJ, Rosenstock J, et al. Tirzepatide versus semaglutide once weekly in patients with type 2 diabetes. N Engl J Med. 2021;385(6):503-515. PubMed
Marso SP, Bain SC, Consoli A, et al. Semaglutide and cardiovascular outcomes in patients with type 2 diabetes. N Engl J Med. 2016;375(19):1834-1844. PubMed
Wilding JPH, Batterham RL, Calanna S, et al. Once-weekly semaglutide in adults with overweight or obesity. N Engl J Med. 2021;384(11):989-1002. PubMed
Mounjaro (tirzepatide) prescribing information. Eli Lilly and Company. FDA Label
Ozempic (semaglutide) prescribing information. Novo Nordisk. FDA Label
American Diabetes Association Professional Practice Committee. Standards of Care in Diabetes, 2023. Diabetes Care. 2023;46(Suppl 1). ADA Standards