Tirzepatide 10 and 15 mg versus semaglutide 2.4 mg in people with obesity or overweight with type 2 diabetes: An indirect treatment comparison

Abstract

Aims

Tirzepatide and semaglutide demonstrated clinically meaningful weight reduction in people with obesity or overweight and type 2 diabetes (T2D) in SURMOUNT‐2 and STEP 2 clinical trials, respectively. In the absence of head‐to‐head trials, this study compared the efficacy of tirzepatide 10 and 15 mg with semaglutide 2.4 mg using an indirect treatment comparison.

Materials and Methods

Mean percent change in weight from baseline, weight reduction ≥5% and mean change in glycated haemoglobin (HbA1c [%]) were compared between tirzepatide 10/15 mg (week 72, SURMOUNT‐2) and semaglutide 2.4 mg (week 68, STEP 2) applying the Bucher method to the efficacy estimand. Sensitivity analyses included the use of matching‐adjusted indirect comparison, treatment regimen estimand and comparing study outcomes at 68 weeks.

Results

Tirzepatide 10 and 15 mg were associated with significantly greater mean percent weight reductions versus semaglutide (mean difference, 10 mg: 2.57%; 15 mg: 4.79%, p < 0.01). Tirzepatide 15 mg had significantly higher odds of achieving ≥5% weight reduction (odds ratio 15 mg: 1.76, 95% CI 1.04–2.97, p = 0.035; odds ratio 10 mg: 1.24, 95% CI 0.75–2.04, p = 0.407), and both tirzepatide doses were associated with significantly greater reductions in HbA1c (%) levels (mean difference, 10 mg: 0.47%; 15 mg: 0.56%, p < 0.001) than semaglutide. Sensitivity analyses were generally consistent with the primary analysis, exceptions including when power was reduced in the matching‐adjusted indirect comparison analyses and in the categorical weight reduction outcome.

Conclusions

This analysis suggested greater reductions in bodyweight and HbA1c (%) levels associated with tirzepatide 10 and 15 mg than with semaglutide 2.4 mg in people with obesity or overweight and T2D.

1. INTRODUCTION

Obesity is a chronic, complex disease involving environmental, social, physiological and other factors that are associated with multiple obesity‐related complications, including type 2 diabetes (T2D), cardiovascular diseases and renal disease. ¹ , ² , ³ In the United States (US), an estimated 42% of adults ≥20 years of age are living with obesity, and approximately 15% of the adult population (≥18 years) have diabetes. ⁴ , ⁵ The overlap between these conditions is substantial, and >90% of people with T2D also live with obesity or overweight. ⁶ Weight management is a critical component of therapy in people with obesity and T2D, as emphasized in the American Association of Clinical Endocrinologists (AACE) obesity guidelines, a consensus report by the American Diabetes Association (ADA), and the European Association for the Study of Diabetes (EASD). ² , ⁷ These guidelines recommend that weight reduction of 5% to ≥15% may substantially lower glycated haemoglobin (HbA1c [%]) levels and provide important health benefits beyond glycaemic control. ² , ⁷ , ⁸ Weight reduction of ≥5% can significantly improve cardiometabolic risk factors associated with obesity ² ; however, greater weight reduction of >10%–15% is often needed for people with some obesity‐related complications. ² , ⁹

Tirzepatide and semaglutide are incretin‐based treatments approved in the US for the treatment of T2D and obesity. ¹⁰ , ¹¹ , ¹² , ¹³ , ¹⁴ , ¹⁵ Tirzepatide is a glucagon‐like peptide‐1 (GLP‐1) and glucose‐dependent insulinotropic polypeptide (GIP) receptor agonist, and semaglutide is a GLP‐1 receptor mono‐agonist. ¹⁶ These obesity medications have demonstrated clinically meaningful weight reduction in people with obesity or overweight and T2D in the phase 3 trials, SURMOUNT‐2 (NCT04657003) and STEP 2 (NCT03552757), respectively. ¹⁷ , ¹⁸ Once weekly tirzepatide 10 and 15 mg resulted in up to 15.7% mean bodyweight reduction (efficacy estimand) at week 72 in the SURMOUNT‐2 trial. ¹⁸ Once weekly semaglutide 2.4 mg resulted in 10.6% bodyweight reduction (trial product estimand) at week 68 in the STEP 2 trial. ¹⁷

While topline results from the head‐to‐head trial of tirzepatide versus semaglutide in people with obesity or overweight without T2D (SURMOUNT‐5, NCT05822830) recently demonstrated that 72 weeks of tirzepatide treatment led to a greater weight reduction compared with semaglutide (20.2% vs. 13.7%), ¹⁹ there are currently no completed head‐to‐head trials comparing tirzepatide and semaglutide for treating obesity in people with T2D. Therefore, indirect treatment comparisons can be useful to provide evidence of the comparative efficacy of treatments. ²⁰ , ²¹ The aim of this study was to compare the efficacy of tirzepatide 10 and 15 mg with semaglutide 2.4 mg (the dose approved for weight management) in people with obesity or overweight and T2D, using an indirect treatment comparison with placebo as the common comparator.

2. METHODS

2.1. Systematic literature review

A systematic literature review (SLR) was conducted to identify relevant randomized controlled trials (RCTs) for inclusion in the indirect treatment comparison network, comparing tirzepatide 10 or 15 mg and semaglutide 2.4 mg. Applying PICOTS criteria (population, intervention, comparator, outcomes, timing, study design; Supplementary Table 1) to the SLR, we identified 2 eligible studies for inclusion: SURMOUNT‐2 (tirzepatide) and STEP 2 (semaglutide). ¹⁷ , ¹⁸

2.2. SURMOUNT‐2 and STEP 2 study and participant characteristics

Table 1 presents the overview of the SURMOUNT‐2 and STEP 2 clinical trials. Both trials were similar in design, inclusion criteria and co‐primary endpoints, aside from the timepoints at which outcomes were assessed (weeks 72 and 68, respectively), although the time at maximum dose was identical (52 weeks) for both tirzepatide 15 mg and semaglutide 2.4 mg.

TABLE 1.

Overview of SURMOUNT‐2 and STEP 2 clinical trials.

Study characteristic	SURMOUNT‐2 18	STEP 2 17
Study design	Double‐blind, randomized, placebo‐controlled, phase 3
Inclusion criteria	Age ≥18 years, BMI ≥27 kg/m2, T2D diagnosis, baseline HbA1c 7%–10%, ≥1 self‐reported unsuccessful dietary effort to lose bodyweight a , b
Time of study conduct	2021–2023	2018–2020
Key regions of enrolment	Argentina, Brazil, India, Japan, Russia, Taiwan, USA	Argentina, Canada, Germany, Greece, India, Japan, Russia, South Africa, Spain, United Arab Emirates, USA, UK
Treatment c	Tirzepatide 10 and 15 mg or placebo for 72 weeks	Semaglutide 1.0 and 2.4 mg or placebo for 68 weeks
Number of randomized participants	938 (1:1:1 ratio)	1210 (1:1:1 ratio)
Number of participants who discontinued study treatment	119 (12.7%)	152 (12.6%)
Number of participants who discontinued the study	79 (8.4%)	46 (3.8%)
Co‐primary endpoint	Percent change in bodyweight from baseline to week 72 and loss of ≥5% of baseline weight at week 72	Percent change in bodyweight from baseline to week 68 and loss of ≥5% of baseline weight at week 68
Key secondary endpoint	Change from baseline in HbA1c (%) levels at week 72	Change from baseline in HbA1c (%) levels at week 68
Estimands	Treatment regimen estimand: average treatment effect at 72 weeks, for all randomly assigned participants, regardless of treatment adherence or use of rescue therapy for hyperglycaemia Efficacy estimand: average treatment effect for all in all randomly assigned participants had they remained on their randomized treatment for the treatment duration (all endpoints), and without using rescue medication for hyperglycaemia (for glycaemic endpoints only)	Treatment policy estimand: average treatment effect at 68 weeks among all randomly assigned patients, regardless of treatment adherence or initiation of rescue therapy Trial product estimand: average treatment effect in all randomly assigned participants, assuming they had remained on treatment for the duration of the trial, and without initiation of obesity rescue medication
Management of gastrointestinal (GI) events (nausea, vomiting or diarrhoea)	Depending on severity, participants were counselled on dietary behaviours, prescribed symptomatic medication (e.g., antiemetic or antidiarrheal medication), allowed temporary study drug interruption for 1 dose, or the study drug was reinitiated at the next lower maintenance dose in a blinded manner. Participants were discontinued from the study drug if all other measures failed	A low starting dose and dose escalation steps were implemented in the trial to mitigate the risk of GI events

Note: Both studies involved an initial dose escalation period and all study participants received lifestyle interventions (e.g., counselling on diet and physical activity). Stable doses of oral anti‐hyperglycaemic medications were allowed with some restrictions (e.g., no GLP‐1 receptor agonists). Complete protocols have been published previously for both studies.

Abbreviations: BMI, body mass index; HbA1c (%), glycated haemoglobin; T2D, type 2 diabetes.

^a Inclusion criterion specific for SURMOUNT‐2: Willing to perform finger stick blood glucose monitoring and self‐ or assisted drug injection.

^b Study‐specific exclusion criteria: SURMOUNT‐2: Have type 1 diabetes mellitus, history of ketosis or hyperosmolar state/coma or any other types of diabetes except T2D; Hypoglycaemia unawareness or severe hypoglycaemia within 6 months of trial; Current or prior treatment with DPP‐4 inhibitors, oral GLP‐1 RA or any injectable therapy for T2D; known clinically significant gastric emptying abnormality or chronic drugs that directly affect GI motility; obesity induced by other endocrinologic disorders or diagnosed monogenetic or syndromic forms of obesity; uncontrolled hypertension; acute or chronic hepatitis, signs and symptoms of any liver disease other than non‐alcoholic fatty liver disease; transplanted organ (corneal transplants allowed) or awaiting an organ transplant; any haematological condition. STEP 2: Use of non‐herbal Chinese medicine or other non‐herbal local medicine with unknown/unspecified content within 90 days before screening; surgery scheduled for the duration of the trial, except for minor surgical procedures.

^c Target dose (following an initial dose escalation period) administered once weekly by subcutaneous injection.

In SURMOUNT‐2, eligible study participants were randomized to receive once weekly tirzepatide 10, 15 mg or a matching placebo for 72 weeks. In STEP 2, participants were randomized to receive once weekly semaglutide 1.0, 2.4 mg or matching placebo for 68 weeks (Table 1). The semaglutide 1.0 mg arm was not included in the current indirect treatment comparison because it is not approved for the treatment of obesity.

2.3. Outcomes

The network diagram for the indirect treatment comparison is presented in Figure 1. The following three outcomes were evaluated:

1. mean percent change in bodyweight from baseline to week 72 (tirzepatide) or week 68 (semaglutide)

2. bodyweight reduction of ≥5% from baseline to week 72 (tirzepatide) or week 68 (semaglutide); and

3. mean change in HbA1c (%) levels from baseline to week 72 (tirzepatide) or week 68 (semaglutide)

FIGURE 1.

Network diagram for the indirect treatment comparison.

2.4. Primary analysis

The primary analysis was conducted using the Bucher method to compare efficacy estimands for the 3 outcomes noted above. ²² , ²³ , ²⁴ Efficacy estimands (terminology used in SURMOUNT‐2) are synonymous with trial product estimands (terminology used in STEP 2). Both represent the average treatment effect relative to placebo for all randomized participants, had they remained on their randomized treatment for the entire planned treatment duration and without using rescue medication for hyperglycaemia (applies to glycaemic endpoint only). Missing values were implicitly handled using a mixed model for repeated measures.

For continuous outcomes (mean percent change in bodyweight and mean change in HbA1c [%]), the relative treatment effect was evaluated using the mean difference. For the binary outcome (bodyweight reduction ≥5%), the relative treatment effect was assessed with the odds ratio (OR). Two‐sided p‐values were considered statistically significant at the level of 0.05, and 95% confidence intervals (CIs) were calculated.

2.5. Sensitivity analyses

Four different sensitivity analyses were conducted to confirm the robustness of the primary analysis (Supplementary Table 2). All outcomes and methodological choices from the primary analysis were kept constant for the sensitivity analyses, except as noted here:

1. The first sensitivity analysis compared treatment regimen estimands (tirzepatide) or treatment policy estimands (semaglutide) which represent the average treatment effect relative to placebo for all randomized participants, regardless of adherence to treatment or use of rescue therapy.

2. The second sensitivity analysis compared efficacy estimands at week 68 for both medications via linear interpolation for tirzepatide to adjust for the difference in overall treatment duration between studies.

3. The third sensitivity analyses used a population‐adjusted approach instead of the Bucher method to compare efficacy estimands. Matching‐adjusted indirect comparison was used for this comparison and included the following variables for matching: ethnicity, sex, baseline HbA1c (%) and treatment with sodium‐glucose co‐transporter 2 (SGLT‐2) inhibitors. ²⁵ These variables were selected based on statistical interactions (standardized mean difference between groups for ethnicity, sex, baseline HbA1c was >0.1) and clinical effect modifiers that were statistically imbalanced (SGLT‐2 inhibitors). Sex was included due to the difference in weight reduction by sex.

4. A fourth post hoc sensitivity analysis was conducted because including the ethnicity variable in the third sensitivity analysis reduced the effective sample size by more than 50%. Using matching‐adjusted indirect comparison, we matched on all participant characteristics in the third sensitivity analysis (sex, baseline HbA1c [%], and treatment with SGLT‐2 inhibitors) except for ethnicity.

3. RESULTS

The primary analysis included 938 SURMOUNT‐2 participants who received tirzepatide 10 mg (n = 312), 15 mg (n = 311) or placebo (n = 315) and 807 STEP 2 participants who received semaglutide 2.4 mg (n = 404) or placebo (n = 403). The effective sample size was the same in the first and second sensitivity analyses that used the Bucher method but was reduced by more than 50% in the third sensitivity analysis (n = 437.4 for SURMOUNT‐2), on account of a large imbalance in the distribution of Hispanic/Latino populations between the studies. In the fourth post hoc sensitivity analysis that leveraged matching‐adjusted indirect comparison, the effective sample size was reduced by less than 3% (n = 917.3 for SURMOUNT‐2) by excluding ethnicity from the matching variables.

Demographic and baseline clinical characteristics were generally similar between the SURMOUNT‐2 and STEP 2 trials, including mean age, sex, duration of T2D, baseline bodyweight, body mass index (BMI) and HbA1c (%) (Table 2 and Supplementary Table 3). However, there were differences between SURMOUNT‐2 and STEP 2 in the proportion of Hispanic/Latino participants (59.8% vs. 12.8%, respectively) and the proportion of participants prescribed SGLT‐2 inhibitors (20.4% vs. 24.8%, respectively; Table 2).

TABLE 2.

Demographic and baseline clinical characteristics of SURMOUNT‐2 ¹⁸ and STEP 2 ¹⁷ participants.

Characteristic a	SURMOUNT‐2 (n = 938)	STEP 2 b (n = 1210)	Standardized mean difference c
Age, years	54.2 (10.6)	55.0 (11.0)	−0.074
Female, n (%)	476 (50.7)	616 (50.9)	−0.004
Bodyweight, kg	100.7 (21.1)	99.8 (21.5)	0.043
BMI, kg/m2	36.1 (6.6)	35.7 (6.3)	0.059
HbA1c, %	8.0 (0.9)	8.1 (0.8)	−0.095
Duration of T2D, years	8.5 (6.5)	8.0 (6.1)	0.087
Ethnicity, n (%)
Hispanic or Latino	561 (59.8)	155 (12.8)	1.12
Use of SGLT‐2 inhibitors, n (%)	191 (20.4)	300 (24.8)	−0.105
Use of sulfonylureas, n (%)	250 (26.7)	308 (25.5)	0.027
Number of anti‐hyperglycaemic medications used, n (%)
0 (diet and physical activity only)	63 (6.7)	56 (4.6)	0.091
1	505 (53.8)	666 (55.0)	−0.024
2	302 (32.2)	398 (32.9)	−0.015
3	65 (6.9)	88 (7.3)	−0.016
≥4	3 (0.3)	2 (0.2)	0.02

Abbreviations: BMI, body mass index; HbA1c (%), glycated haemoglobin; ITC, indirect treatment comparison; SD, standard deviation; SGLT‐2, sodium‐glucose co‐transporter 2; T2D, type 2 diabetes.

^a Data presented as mean (SD) unless otherwise noted.

^b Data from STEP 2 includes the semaglutide 1.0 mg treatment arm, which was not included in this ITC.

^c A standardized mean difference of ≥0.1 is considered to be a relevant difference.

3.1. Mean percent change in bodyweight

Tirzepatide 10 and 15 mg doses were associated with significantly greater reductions in mean percent change in bodyweight at week 72 compared with semaglutide 2.4 mg at week 68 (Figure 2A). The mean difference was −2.57 (95% CI: −4.24, −0.90; p = 0.003) for tirzepatide 10 mg and −4.79 (95% CI: −6.47, −3.12; p < 0.001) for tirzepatide 15 mg versus semaglutide 2.4 mg.

FIGURE 2.

Comparison of tirzepatide versus semaglutide for mean percent change from baseline in bodyweight and ≥5% bodyweight reduction from baseline. Indirect treatment comparison of tirzepatide 10 and 15 mg versus semaglutide 2.4 mg for (A) mean percent change from baseline in bodyweight, and (B) odds of participants achieving ≥5% bodyweight reduction from baseline (primary and sensitivity analyses; see Supplementary Table 2 for details). All outcomes and methodological choices from the primary analysis were kept constant for the sensitivity analyses except: Sensitivity analysis 1: comparison of treatment regimen estimand (tirzepatide) or treatment policy estimand (semaglutide) Sensitivity analysis 2: comparing endpoints at 68 weeks for both medications. Sensitivity analysis 3: matching‐adjusted indirect comparison adjusting for ethnicity, sex, baseline HbA1c (%) and treatment with sodium‐glucose co‐transporter 2 inhibitors. Sensitivity analysis 4: matching‐adjusted indirect comparison adjusting for sex, baseline HbA1c and treatment with sodium‐glucose co‐transporter 2 inhibitors. Mean difference <0 favours tirzepatide (panel A); OR >1 favours tirzepatide (panel B). CI, confidence interval; OR, odds ratio.

The results and direction of all sensitivity analyses were consistent with primary analysis results, with one exception. In the third sensitivity analysis, the matching‐adjusted indirect comparison for the difference in mean weight reduction for tirzepatide 10 mg versus semaglutide 2.4 mg showed numerically greater weight reduction for tirzepatide that was not statistically significant (mean difference − 1.36; 95% CI −3.27, 0.55; p = 0.162; Figure 2A).

3.2. Achievement of ≥5% bodyweight reduction

The odds of participants achieving ≥5% reduction in bodyweight were significantly higher for tirzepatide 15 mg at week 72 than for semaglutide 2.4 mg at week 68 (OR 1.76; 95% CI: 1.04, 2.97; p = 0.035). The odds of tirzepatide 10 mg over semaglutide 2.4 mg were not statistically significant for achieving ≥5% bodyweight reduction in the primary analysis (OR 1.24; 95% CI: 0.75, 2.04; p = 0.407), although the direction of results favoured tirzepatide 10 mg (Figure 2B).

Results of sensitivity analyses 1 and 2 (i.e., comparison of treatment regimen estimands and comparison of outcomes at week 68 for both medications, respectively) also demonstrated that the odds of achieving ≥5% bodyweight reduction were significantly higher for tirzepatide 15 mg than semaglutide 2.4 mg (p < 0.01). No significant differences between treatment groups were found in sensitivity analyses 3 and 4 (i.e., using the matching‐adjusted indirect comparison method), though the direction of results favoured tirzepatide 15 mg (Figure 2B).

3.3. Mean change in HbA1c (%) levels

The primary analysis showed that treatment with tirzepatide 10 and 15 mg doses was associated with significantly greater reductions in HbA1c (%) compared with semaglutide 2.4 mg (Figure 3). The mean difference was −0.47% (95% CI: −0.70, −0.24; p < 0.001) for tirzepatide 10 mg and −0.56% (95% CI: −0.79, −0.33; p < 0.001) for tirzepatide 15 mg versus semaglutide 2.4 mg. The results and direction of all 4 sensitivity analyses were consistent with primary analysis results (Figure 3).

FIGURE 3.

Comparison of tirzepatide versus semaglutide for mean change in HbA1c (%). Indirect treatment comparison of tirzepatide 10 and 15 mg versus semaglutide 2.4 mg for mean change in HbA1c (primary and sensitivity analyses; see Supplementary Table 2 for details). All outcomes and methodological choices from the primary analysis were kept constant for the sensitivity analyses except: Sensitivity analysis 1: comparison of treatment regimen estimand (tirzepatide) or treatment policy estimand (semaglutide). Sensitivity analysis 2: comparing endpoints at 68 weeks for both medications. Sensitivity analysis 3: matching‐adjusted indirect comparison adjusting for ethnicity, sex, baseline HbA1c (%) and treatment with sodium‐glucose co‐transporter 2 inhibitors. Sensitivity analysis 4: matching‐adjusted indirect comparison adjusting for sex, baseline HbA1c and treatment with sodium‐glucose co‐transporter 2 inhibitors. Mean difference <0 favours tirzepatide. CI, confidence interval; HbA1c (%), glycated haemoglobin.

4. DISCUSSION

This indirect treatment comparison used data from the 72‐week SURMOUNT‐2 trial and the 68‐week STEP 2 trial in people with obesity or overweight with T2D. In the absence of head‐to‐head comparative trials between tirzepatide and semaglutide in people with obesity or overweight with T2D, this analysis provides indirect evidence that tirzepatide treatment (10 and 15 mg) may have been associated with greater weight and HbA1c reduction than semaglutide (2.4 mg). The findings are consistent with the recent topline results from the SURMOUNT‐5 study in people with obesity or overweight without T2D ¹⁹ and provide clinically relevant information for healthcare professionals and other decision‐makers regarding the use of pharmacotherapy to improve the delivery of care and achieve targets for weight reduction and HbA1c control in people with obesity or overweight with T2D.

Our findings are consistent with a previous indirect treatment comparison (SURMOUNT‐1 vs. STEP 1) showing 5.9% greater weight reduction for tirzepatide 15 mg versus semaglutide 2.4 mg in people with overweight or obesity without T2D ²⁶ and with topline results of a head‐to‐head trial of tirzepatide versus semaglutide in people with obesity or overweight without T2D (SURMOUNT‐5, NCT05822830), which demonstrated that 72 weeks of tirzepatide treatment (10 or 15 mg) led to greater weight reduction compared with semaglutide (1.7 or 2.4 mg; mean weight reduction: 20.2% for tirzepatide, 13.7% for semaglutide). ¹⁹ Inclusion of people with T2D in the current analysis likely explains the lower weight reduction difference observed here versus the previous work. In addition, the phase 3 SURPASS‐2 trial, which compared efficacy and safety of tirzepatide and semaglutide among people with T2D, with or without obesity, reported a mean difference in bodyweight change from baseline at week 40 between tirzepatide (15 mg) and semaglutide (1 mg) of −6.2 kg (efficacy estimand). ²⁷ Findings of the current analysis are consistent with SURPASS‐2 results, given the differences in the semaglutide dose (2.4 vs. 1 mg), treatment duration (72/68 vs. 40 weeks) and study populations (people with obesity/overweight and T2D vs. people with T2D with or without obesity), respectively.

According to AACE, ADA and EASD clinical guidelines, modest and sustained weight reduction (5%–10%) in people living with obesity and T2D may treat or prevent the development of multiple obesity‐related complications. ² , ⁷ , ²⁸ The results from this indirect treatment comparison showed a mean difference in percent bodyweight reduction between tirzepatide 15 mg and semaglutide 2.4 mg ranging from 4.4% to 5.3%, with up to 2.2 times greater odds of achieving ≥5% weight reduction. Importantly, research has shown that the degree of weight reduction to achieve clinically meaningful benefits varies as a function of obesity‐related complications. In people with obesity and T2D, weight reduction of 10%–15% or more is associated with progressive improvements in HbA1c levels, obstructive sleep apnea, metabolic dysfunction‐associated steatotic liver disease and prevention of cardiovascular events. ² Furthermore, 5% bodyweight reduction leads to significant improvements in metabolic function, including improved insulin sensitivity and β cell function, with additional weight reduction contributing to changes in adipose tissue mass, lipid metabolism and oxidative stress. ²⁹ , ³⁰

The SURMOUNT‐2 and STEP 2 trials demonstrated weight and HbA1c reduction following tirzepatide and semaglutide treatment, with most people achieving recommended HbA1c targets. ¹⁷ , ¹⁸ Reaching recommended targets for HbA1c is also critical for preventing microvascular disease complications in T2D. The mean difference in HbA1c levels between tirzepatide and semaglutide in this indirect treatment comparison ranged from 0.35 to 0.67, and physicians consider 0.5% a clinically meaningful reduction in a real‐world setting. ³¹ Our findings align with the estimated treatment differences in HbA1c levels reported in SURPASS‐2 (−0.15, −0.39 and −0.45 for tirzepatide 5, 10 or 15 mg, respectively, vs. semaglutide 1 mg). ²⁷ It is possible that the greater reductions in bodyweight and HbA1c levels favouring tirzepatide over semaglutide observed in the current indirect treatment comparison could translate to clinical benefits for both obesity and T2D, although this will require further research.

A potential mechanism for the greater weight reduction associated with tirzepatide versus semaglutide treatment may be that while both drugs are GLP‐1 receptor agonists, tirzepatide also activates GIP receptors. ¹¹ , ¹⁴ GLP‐1 acts both centrally and peripherally to regulate body weight and GIP acts centrally to regulate body weight by modulating appetite and food intake. ³² Although the specific properties that confer differences in efficacy are unclear, the different mechanisms of action are a possible driver that can be explored in future research.

Including ethnicity as a matching variable in the third sensitivity analysis decreased the effective sample size by over 50%, due to a higher number of Hispanic and Latino participants in SURMOUNT‐2 compared to STEP 2. Consequently, a fourth post hoc sensitivity analysis was performed, omitting ethnicity as a matching variable. In a subgroup analysis of the SURMOUNT‐2 trial, Frias et al. demonstrated that Hispanic and Latino participants experienced less weight reduction than their non‐Hispanic or Latino counterparts, ²⁹ which would be expected to result in smaller weight changes with tirzepatide than with semaglutide in the current analysis. Nevertheless, results from the fourth sensitivity analysis demonstrated significantly greater reductions in mean percent change in bodyweight with tirzepatide compared to semaglutide, suggesting that ethnicity is unlikely to significantly influence the favorable outcomes observed with tirzepatide treatment relative to semaglutide.

Strengths of our analysis included an SLR to identify eligible studies for inclusion in the indirect treatment comparison, as well as undertaking several sensitivity analyses using different estimands, time points for outcome evaluation, and 2 different comparison methods – Bucher and matching‐adjusted indirect comparison. The indirect treatment comparison also evaluated 3 different outcomes, although future analyses could be considered for additional outcomes, such as higher weight reduction thresholds of ≥10% and ≥15% of bodyweight and ORs for achieving HbA1c targets of <7% and ≤6.5%.

The study also had some limitations that should be considered when interpreting these results. Specifically, these limitations include the absence of tolerability/safety analyses, which are often not feasible when using indirect treatment comparison techniques due to methodological differences between trials, such as definitions of adverse events and serious adverse events, differences in the methods of safety data collection and different study durations (72 vs. 68 weeks). Next, only 2 studies were eligible for inclusion in the indirect treatment comparison, which may limit the robustness of the comparison. Furthermore, differences in study design or endpoint reporting methods of SURMOUNT‐2 and STEP 2, including measured and unmeasured baseline variables and the implementation of the lifestyle modification in the trials, may have biased the results of this analysis. Notably, the placebo groups had 3.2% and 3.4% mean bodyweight reduction in SURMOUNT‐2 and STEP 2, respectively, suggesting a degree of similarity in how the lifestyle interventions were operationalized between the trials. The findings of our analysis need to be verified in RCTs because indirect treatment comparisons may not account for residual confounding arising from unobserved differences between trials.

In conclusion, results of this indirect treatment comparison suggested greater reductions in bodyweight and HbA1c levels associated with tirzepatide 10 and 15 mg than with semaglutide 2.4 mg in people with obesity or overweight and T2D. In the absence of head‐to‐head clinical trials, the results of this analysis may potentially guide clinical and pharmacoeconomic decisions in treating people with obesity or overweight and T2D.

AUTHOR CONTRIBUTIONS

Design: Emily R. Hankosky, Raleigh Malik, Xuanyao He. Conduct/data collection: Emily R. Hankosky, Xuanyao He, Fangyu Wang, Xiaotian Michelle Zhang, Anthony Niemeyer, Julia Fraseur Brumm, W. Timothy Garvey. Analysis: Xuanyao He, Fangyu Wang, W. Timothy Garvey. Interpretation: Emily R. Hankosky, Xuanyao He, Fangyu Wang, Xiaotian Michelle Zhang, Anthony Niemeyer, Julia Fraseur Brumm, W. Timothy Garvey. Writing manuscript: All authors.

FUNDING INFORMATION

The study was funded by Eli Lilly and Company, Indianapolis, United States.

CONFLICT OF INTEREST STATEMENT

Emily R. Hankosky, Xuanyao He, Julia Fraseur Brumm, Fangyu Wang, Anthony Niemeyer and Xiaotian Michelle Zhang are employees and stockholders of Eli Lilly and Company, Indianapolis, United States. Raleigh Malik was an employee of Eli Lilly and Company at the time of the study. W. Timothy Garvey has served as a consultant on advisory boards for Boehringer Ingelheim, Eli Lilly and Company, Novo Nordisk, Pfizer, Fractyl Health, Alnylam Pharmaceuticals, Inogen, Zealand, Carmot/Roche, Regeneron and Merck, and as a site principal investigator for multi‐centred clinical trials sponsored by his university and funded by Novo Nordisk, Eli Lilly and Company, Epitomee, Neurovalens and Pfizer.

PEER REVIEW

The peer review history for this article is available at https://www.webofscience.com/api/gateway/wos/peer‐review/10.1111/dom.16401.

Supporting information

Data S1. Supporting information.

Hankosky ER, He X, Malik R, et al. Tirzepatide 10 and 15 mg versus semaglutide 2.4 mg in people with obesity or overweight with type 2 diabetes: An indirect treatment comparison. Diabetes Obes Metab. 2025;27(7):3757‐3765. doi: 10.1111/dom.16401

DATA AVAILABILITY STATEMENT

Data are available from the corresponding author upon reasonable request.