Effect of semaglutide versus placebo on cardiorenal outcomes by prior cardiovascular disease and baseline body mass index: Pooled post hoc analysis of SUSTAIN 6 and PIONEER 6

Abstract

Aims

Cardiorenal effects of semaglutide in people with type 2 diabetes (T2D) at high cardiovascular (CV) risk were investigated.

Materials and Methods

Post hoc analyses of pooled SUSTAIN 6 (NCT01720446) and PIONEER 6 (NCT02692716) data assessed time to primary major adverse CV events (MACE; CV death, non‐fatal myocardial infarction, or non‐fatal stroke), expanded MACE (MACE + hospitalisation for unstable angina or heart failure), CV death, all‐cause death, and new or worsening nephropathy. The impact of body weight (BW) changes on primary MACE risk was also evaluated. Participants were stratified by prior CV disease (CVD) status and baseline body mass index (BMI).

Results

Semaglutide significantly reduced the risk of primary and expanded MACE, with a nonsignificant risk reduction of CV and all‐cause death versus placebo in the overall population; the effect was consistent across all subgroups (p _INT >0.05 for all comparisons). Semaglutide consistently reduced nephropathy risk versus placebo in the SUSTAIN 6 population (HR [95% CI]: 0.64 [0.46; 0.88], p = 0.0054) and across all subgroups (p _INT >0.05 for all comparisons). When accounting for BW changes, treatment effects on primary MACE risk in the overall population and by BMI subgroups remained similar compared with the results of the main analysis.

Conclusions

Semaglutide treatment improved cardiorenal outcomes versus placebo in people with T2D, regardless of prior CVD and baseline BMI. This improvement was observed even when accounting for changes in BW, indicating direct effects of semaglutide on the cardiorenal system. This analysis supports the broad efficacy of semaglutide in a diverse T2D population.

1. INTRODUCTION

People with prediabetes or type 2 diabetes (T2D) have an increased risk of cardiorenal events, ¹ , ² , ³ , ⁴ emphasising the link between cardiovascular (CV) disease (CVD), chronic kidney disease (CKD), and suboptimal glycaemic control. ⁵ T2D is a complex disease with a high degree of heterogeneity ⁶ , ⁷ , ⁸ and several underlying pathological mechanisms that affect the risk of comorbidities and response to oral glucose‐lowering drugs. ⁶ , ⁷ The pathogenesis of T2D in people with or without overweight/obesity also differs. ⁸

Glucagon‐like peptide‐1 receptor agonists (GLP‐1RAs) have multifactorial effects in people with T2D because this drug class targets the pancreas, brain, and insulin‐sensitive tissues to regulate glycaemic control and food intake, ⁹ , ¹⁰ also resulting in body weight (BW) reductions. ¹¹ Furthermore, GLP‐1RAs affect the CV system ¹² and kidneys. ¹³ A meta‐analysis of CV outcomes trials (CVOTs) has shown that GLP‐1RAs reduced the risk of major adverse CV events (MACE) and worsening of kidney function by 14% and 18%, respectively, in people with T2D and with, or at high/very high risk of, CVD. ¹⁴ Specifically, the SUSTAIN 6 (Trial to Evaluate Cardiovascular and Other Long‐term Outcomes with Semaglutide in Subjects with Type 2 Diabetes) trial has shown CV benefits of the once‐weekly (OW) subcutaneous (s.c.) GLP‐1RA semaglutide versus placebo in people with T2D at high CV risk or with established CVD, ¹⁵ whereas once‐daily (OD) oral semaglutide showed noninferiority on the risk of MACE versus placebo in the PIONEER 6 (A Trial Investigating the Cardiovascular Safety of Oral Semaglutide in Subjects With Type 2 Diabetes) trial. ¹⁶ In a pooled analysis of the SUSTAIN 6 and PIONEER 6 CVOTs, semaglutide reduced the risk of primary MACE (i.e., a composite outcome of CV death, non‐fatal myocardial infarction [MI], or non‐fatal stroke) versus placebo, independent of baseline CV risk, with the exception of people with prior heart failure (HF). ¹⁷ In an analysis of the LEADER (Liraglutide Effect and Action in Diabetes: Evaluation of Cardiovascular Outcome Results) and SUSTAIN 6 trials, the risk reduction of primary MACE, expanded MACE, CV death, and nephropathy with OD s.c. liraglutide or OW s.c. semaglutide versus placebo was generally consistent across subpopulations with different baseline body mass index (BMI) ranges. ¹⁸ In addition, the kidney outcomes trial FLOW (A Research Study to See How Semaglutide Works Compared with Placebo in People With Type 2 Diabetes and Chronic Kidney Disease; NCT03819153) demonstrated that OW s.c. semaglutide 1.0 mg reduced the risk of major kidney disease events, MACE, and all‐cause death by 24%, 18%, and 20%, respectively, versus placebo in people with T2D and CKD. ¹⁹

The mechanism of action of semaglutide in CVD and CKD remains unclear. Although it has been considered that GLP‐1RA‐induced reductions in BW might indirectly contribute to cardiorenal benefits, ¹³ , ²⁰ , ²¹ previous analyses have shown that changes in BW did not impact the effects of GLP‐1RAs on CV and kidney‐related outcomes. ²² , ²³ , ²⁴

As CV risk factors and baseline BMI variation might affect drug responses in people with T2D, ⁶ the objective of this post hoc analysis was to explore whether the effect of semaglutide versus placebo on a broad range of CV outcomes and a kidney composite outcome is affected by prior CVD or differences in baseline BMI, and to evaluate whether changes in BW account for the effect of semaglutide on primary MACE risk.

To our knowledge, this is the first analysis that explores the impact of CVD history in further detail, such as previous history of revascularisation, MI or stroke/transient ischaemic attack (TIA), and that investigates the cardiorenal effects of semaglutide across a wider range of BMI categories, also including people with lean BW, to gain further insights on whether semaglutide exerts its beneficial effects irrespective of weight loss. This information might contribute to the use of semaglutide in real‐world clinical practice.

2. METHODS

2.1. Study design and population

Individual participant data from participants in the SUSTAIN 6 (N = 3297) and PIONEER 6 (N = 3183) CVOTs, receiving semaglutide (SUSTAIN 6: OW s.c. 0.5 or 1.0 mg; PIONEER 6: OD oral 14 mg) or placebo, were used in this analysis. Data from the s.c. and oral semaglutide arms were pooled and compared with pooled placebo data from both trials. The SUSTAIN 6 and PIONEER 6 CVOTs have been described previously ¹⁵ , ¹⁶ ; however, relevant information is provided here for clarity.

The SUSTAIN 6 and PIONEER 6 CVOTs were event‐driven trials, with a minimum of 122 MACE events needed to ascertain statistical power. ¹⁵ , ¹⁶ A minimum trial duration of 109 weeks was defined for SUSTAIN 6, ¹⁵ whereas no minimal trial duration was established for PIONEER 6. ¹⁶ The median study observation time for the SUSTAIN 6 and PIONEER 6 trials was 2.1 years and 15.9 months, respectively. ¹⁵ , ¹⁶ Key inclusion criteria for the participants of both CVOTs were an age of ≥50 years and established CVD (including previous CV, cerebrovascular or peripheral vascular disease), chronic HF or CKD, or an age of ≥60 years and at least one CV risk factor. ¹⁵ , ¹⁶ In both CVOTs, key kidney‐related exclusion criteria were long‐term or intermittent haemodialysis or peritoneal dialysis ¹⁵ , ¹⁶ ; for PIONEER 6, an estimated glomerular filtration rate (eGFR) <30 mL/min/1.73 m² was also an exclusion criterion. ¹⁶

2.2. Subgroups by prior CVD or baseline BMI

The overall pooled study population was divided into subgroups based on the presence of CVD (coronary, carotid or peripheral arterial revascularisation, MI, and stroke/TIA) before randomisation. Participants were also divided into subgroups based on BMI ranges at baseline: <25 kg/m² (indicating lean BW), ≥25 to <30kg/m² (indicating overweight), and ≥30kg/m² (indicating obesity). ²⁵

2.3. Study outcomes

Both SUSTAIN 6 and PIONEER 6 used identical definitions for primary MACE and expanded MACE endpoints. ¹⁵ , ¹⁶ Time to first primary MACE and its individual components, expanded MACE, CV death, and all‐cause death, with semaglutide versus placebo was evaluated in the overall study population and by prior CVD or baseline BMI subgroups and by continuous BMI values at baseline. Primary MACE was defined as a composite of CV death, non‐fatal MI, or non‐fatal stroke, while expanded MACE was defined as components of primary MACE and hospitalisation for unstable angina pectoris or HF. Hospitalisation for HF was defined as an event including hospital admission with a primary HF diagnosis, a length of stay for ≥24 h (or over a change in calendar date if admission and discharge times were unavailable), documented new or worsening of ≥1 symptom due to HF (including dyspnoea, decreased exercise tolerance, fatigue, and other symptoms of end‐organ perfusion or volume overload) on presentation, evidence of new or worsening HF consisting of at least two physical examinations or one physical examination and at least one laboratory criterion, and initiation or intensification of at least one treatment for HF. ¹⁵ , ¹⁶ All outcomes were adjudicated in a double‐blinded fashion by an external, independent event‐adjudication committee. ¹⁵ , ¹⁶

Time to first new or worsening of nephropathy with semaglutide versus placebo was evaluated in the SUSTAIN 6 population only, as kidney‐related outcomes were not assessed in PIONEER 6, ¹⁶ and by prior CVD or baseline BMI subgroups and by continuous BMI values at baseline within this population. The kidney composite outcome was defined as time to first occurrence of new onset of persistent macroalbuminuria (≥300 mg/g), persistent doubling of serum creatinine level and eGFR ≤45 mL/min/1.73 m², continuous kidney replacement therapy, or death due to kidney disease.

An exploratory outcome evaluating whether changes in BW during the SUSTAIN 6 and PIONEER 6 CVOTs also accounted for the effect of semaglutide on primary MACE risk was included.

The absolute and percentual mean change in BW from baseline to first event of primary MACE or censoring, defined as loss to follow‐up, death, or end of study, was analysed in the overall study population and by baseline BMI subgroups. The relative risk of primary MACE was assessed in relation to BW change or treatment effect.

2.4. Statistical analysis

Baseline characteristics were compared between the semaglutide and placebo arms in the overall study population and between the semaglutide and placebo arms within each prior CVD or baseline BMI subgroup, using the two‐sample t test for continuous variables and the chi‐square test for categorical variables. As the kidney composite outcome was assessed in the SUSTAIN 6 population only, mean eGFR at baseline was analysed using both pooled trial data and SUSTAIN 6 data.

Time from randomisation to first event of CV outcomes was analysed in the overall study population and by subgroups. A Cox proportional hazards model was applied, and data are presented as hazard ratios (HR) with 95% confidence intervals (CIs). The proportional hazards assumption was tested using Schoenfeld residuals. The global Schoenfeld residuals correlation test confirmed the validity of the assumption (p values >0.05 for all predictors). Time from randomisation to non‐fatal stroke was analysed in the subgroups with BMI <30 kg/m² and ≥30 kg/m² only, as there were no events recorded for those with a BMI <25 kg/m² who were treated with semaglutide. The model was applied to analyse time from randomisation to first event of new or worsening nephropathy in the SUSTAIN 6 population and by respective subgroups. For both analyses evaluating CV or kidney‐related outcomes, treatment, subgroup, and interaction of treatment by subgroup were used as categorical fixed factors, and the model was stratified by trial (SUSTAIN 6, PIONEER 6).

Time from randomisation to first event of CV outcomes, as well as new or worsening nephropathy in the SUSTAIN 6 population, was analysed by baseline BMI as a continuous variable. A Cox proportional hazards model was applied, using treatment, BMI (continuous variable), and interaction of treatment by BMI (continuous variable) as fixed factors; the model was stratified by trial (SUSTAIN 6, PIONEER 6). The mean BMI value for each of the abovementioned BMI categories was calculated to identify intercepts on the BMI continuum of the overall study population and the SUSTAIN 6 population. The mean BMI values were then used to estimate HRs with 95% CIs with the Cox proportional hazards model.

The Cox proportional hazards models were adjusted for covariates that were significantly different between semaglutide and placebo arms at baseline. No adjustment for multiplicity was performed. An interaction p value (p _INT) of <0.05 was considered significant, indicating that the effect of semaglutide versus placebo between prior CVD or baseline BMI subgroups and baseline BMI values was inconsistent.

To account for variations in weight loss during the SUSTAIN 6 and PIONEER 6 CVOTs on the effect of semaglutide on primary MACE risk across baseline BMI subgroups, a Cox proportional hazards model was used with BW change as a time‐dependent variable. In this model, treatment was used as a fixed factor, baseline BW was used as a covariate (for the overall study population only), and the model was stratified by trial (SUSTAIN 6, PIONEER 6) and presence of CVD at baseline (yes/no). Data are presented as HR with 95% CIs.

Kaplan–Meier (KM) analyses were conducted for all endpoints and subgroups.

3. RESULTS

3.1. Baseline characteristics

Baseline characteristics for the overall study population (N = 6480) and the overall study population by prior CVD and baseline BMI are shown in Appendices A–E; SUSTAIN 6 and PIONEER 6 baseline data are published, ¹⁵ , ¹⁶ and baseline characteristics are comparable across both studies; baseline characteristics in the pooled population were generally balanced across subgroups.

Mean age was 64.4–65.9 years, ≥53.0% of participants were male, participants had a suboptimal glycaemic control (glycated haemoglobin [HbA1c] 8.3%–8.7%) at baseline, and the mean duration of T2D was 13.9–16.2 years. Across all subgroups, 49.3%–62.9% of participants used insulin at baseline. Furthermore, >80% of participants in the overall study population had experienced at least one CVD event at baseline (Appendix A). The mean eGFR was 73.6–77.5 mL/min/1.73 m² across all subgroups, and mean eGFR was similar between treatment arms in the overall study population and the SUSTAIN 6 population at baseline (Appendices B–E).

The mean HbA1c at baseline was mostly similar between the semaglutide and placebo arms within each prior CVD or baseline BMI subgroup; however, participants in the semaglutide arm with prior stroke/TIA had higher HbA1c at baseline than those in the placebo arm (Appendix D), and analyses for all outcomes in this subgroup were adjusted for HbA1c.

3.2. The effect of semaglutide on cardiorenal outcomes in the overall study population

A previous evaluation of the overall SUSTAIN 6 and PIONEER 6 population ¹⁷ showed a significantly reduced risk of primary MACE (HR [95% CI] 0.76 [0.62; 0.92], p = 0.0062) and non‐fatal stroke (HR [95% CI] 0.65 [0.43; 0.97], p = 0.0346) with semaglutide versus placebo, whereas there were no significant differences in the risk of CV death (HR [95% CI] 0.78 [0.56; 1.10], p = 0.1553) and non‐fatal MI (HR [95% CI] 0.88 [0.66; 1.18], p = 0.4000) between semaglutide and placebo. Similarly, this analysis shows the risk of expanded MACE (HR [95% CI] 0.80 [0.68; 0.95], p = 0.0097) was significantly reduced with semaglutide versus placebo in the overall study population, whereas there was a nonsignificant risk reduction of all‐cause death (HR [95% CI] 0.81 [0.61; 1.08], p = 0.1533).

Analysis of the overall SUSTAIN 6 population showed that the risk of new or worsening nephropathy (HR [95% CI] 0.64 [0.46; 0.88], p = 0.0054)D was reduced with semaglutide versus placebo. ¹⁵

3.3. The effect of semaglutide on cardiorenal outcomes by prior CVD

Irrespective of prior CVD status, the risk of CV outcomes was consistently lower with semaglutide versus placebo (p _INT >0.05 for all comparisons; Figures 1, 2, 3). Moreover, semaglutide consistently reduced the risk of new or worsening nephropathy versus placebo across subgroups with or without prior CVD (p _INT >0.05 for all comparisons; Figures 1, 2, 3).

FIGURE 1.

The effect of semaglutide versus placebo on cardiorenal outcomes by prior revascularisation^†. Time from randomisation to first event was analysed using a Cox proportional hazards model with treatment, subgroup, and interaction of treatment by subgroup as categorical fixed factors; the model was stratified by trial. A p _INT <0.05 was considered significant, indicating that the effect of semaglutide versus placebo between subgroups was inconsistent. ^†Including coronary, carotid or peripheral arterial revascularisation. ^‡Primary MACE was defined as composite of CV death, non‐fatal MI, or non‐fatal stroke. ^§Expanded MACE was defined as components of primary MACE and hospitalisation for unstable angina pectoris or heart failure. ^¶Data from SUSTAIN 6 population only (N = 3297). Defined as time to first occurrence of new onset of persistent macroalbuminuria (≥300 mg/g), persistent doubling of serum creatinine level and eGFR ≤45 mL/min/1.73 m², continuous kidney replacement therapy, or death due to kidney disease. %, proportion of participants with event; CI, confidence interval; CV, cardiovascular; eGFR, estimated glomerular filtration rate; HR, hazard ratio; MACE, major adverse cardiovascular events; MI, myocardial infarction; N, number of participants in baseline category; n, number of participants with event; p _INT, interaction p value.

FIGURE 2.

The effect of semaglutide versus placebo on cardiorenal outcomes by prior MI. Time from randomisation to first event was analysed using a Cox proportional hazards model with treatment, subgroup, and interaction of treatment by subgroup as categorical fixed factors; the model was stratified by trial. A p _INT <0.05 was considered significant, indicating that the effect of semaglutide versus placebo between subgroups was inconsistent. ^†Primary MACE was defined as composite of CV death, non‐fatal MI, or non‐fatal stroke. ^‡Expanded MACE was defined as components of primary MACE and hospitalisation for unstable angina pectoris or heart failure. ^§Data from SUSTAIN 6 population only (N = 3297). Defined as time to first occurrence of new onset of persistent macroalbuminuria (≥300 mg/g), persistent doubling of serum creatinine level and eGFR ≤45 mL/min/1.73 m², continuous kidney replacement therapy, or death due to kidney disease. %, proportion of participants with event; CI, confidence interval; CV, cardiovascular; eGFR, estimated glomerular filtration rate; HR, hazard ratio; MACE, major adverse cardiovascular events; MI, myocardial infarction; N, number of participants in baseline category; n, number of participants with event; p _INT, interaction p value.

FIGURE 3.

The effect of semaglutide versus placebo on cardiorenal outcomes by prior stroke/TIA. Time from randomisation to first event was analysed using a Cox proportional hazards model with treatment, subgroup, and interaction of treatment by subgroup as categorical fixed factors; the model was stratified by trial. A p _INT <0.05 was considered significant, indicating that the effect of semaglutide versus placebo between subgroups was inconsistent. ^†Primary MACE was defined as composite of CV death, non‐fatal MI, or non‐fatal stroke. ^‡Expanded MACE was defined as components of primary MACE and hospitalisation for unstable angina pectoris or heart failure. ^§Data from SUSTAIN 6 population only (N = 3297). Defined as time to first occurrence of new onset of persistent macroalbuminuria (≥300 mg/g), persistent doubling of serum creatinine level and eGFR ≤45 mL/min/1.73 m², continuous kidney replacement therapy, or death due to kidney disease. %, proportion of participants with event; CI, confidence interval; CV, cardiovascular; eGFR, estimated glomerular filtration rate; HR, hazard ratio; MACE, major adverse cardiovascular events; MI, myocardial infarction; N, number of participants in baseline category; n, number of participants with event; p _INT, interaction p value; TIA, transient ischaemic attack.

Results from the analysis adjusted for the difference in baseline HbA1c between treatment arms in participants with prior stroke/TIA (Appendix D) were consistent with the unadjusted analysis, demonstrating the robustness of the data (Appendix F).

KM plots showing primary MACE over time (Appendix G) by prior revascularisation, MI, and stroke/TIA demonstrated a lower number of events for semaglutide compared with placebo across all subgroups.

3.4. The effect of semaglutide on cardiorenal outcomes by baseline BMI

The risk of CV outcomes was consistently reduced with semaglutide versus placebo across baseline BMI subgroups (p _INT >0.05 for all comparisons; Figure 4). Regarding the kidney composite outcome, semaglutide reduced the risk of new or worsening nephropathy versus placebo regardless of baseline BMI (p _INT >0.05 for all comparisons; Figure 4).

FIGURE 4.

The effect of semaglutide versus placebo on cardiorenal outcomes by baseline BMI subgroups. Time from randomisation to first event was analysed using a Cox proportional hazards model with treatment, subgroup, and interaction of treatment by subgroup as categorical fixed factors; the model was stratified by trial. A p _INT <0.05 was considered significant, indicating that the effect of semaglutide versus placebo between subgroups was inconsistent. Participants without baseline BMI (n = 7 and n = 1 from SUSTAIN 6 and PIONEER 6, respectively) were excluded from the analysis. ^†Primary MACE was defined as composite of CV death, non‐fatal MI, or non‐fatal stroke. ^‡Expanded MACE was defined as components of primary MACE and hospitalisation for unstable angina pectoris or heart failure. ^§Time from randomisation to non‐fatal stroke outcome was analysed in the subgroups with BMI <30 kg/m² and ≥30 kg/m² only, as there were no events for those with a BMI <25 kg/m² who were treated with semaglutide (i.e., data from the BMI <25 and ≥25 to <30 kg/m² subgroups were pooled). ^¶Data from SUSTAIN 6 population only (N = 3297). Defined as time to first occurrence of new onset of persistent macroalbuminuria (≥300 mg/g), persistent doubling of serum creatinine level and eGFR ≤45 mL/min/1.73 m², continuous kidney replacement therapy, or death due to kidney disease. %, proportion of participants with event; BMI, body mass index; CI, confidence interval; CV, cardiovascular; eGFR, estimated glomerular filtration rate; HR, hazard ratio; MACE, major adverse cardiovascular events; MI, myocardial infarction; N, number of participants in baseline category; n, number of participants with event; p _INT, interaction p value.

Results for CV outcomes when using baseline BMI as a continuous variable were consistent compared with using BMI subgroups. Despite a statistically significant interaction (p _INT = 0.037) between baseline BMI and treatment on the risk of new or worsening nephropathy, beneficial effects of semaglutide persisted at various BMI values (Appendix H).

KM plots showing primary MACE over time (Appendix I) by BMI subgroup demonstrated a lower number of events for semaglutide compared with placebo across all subgroups.

3.5. Accounting for BW changes from baseline on primary MACE risk

The reduction in absolute mean BW from baseline to time to first primary MACE or censoring was −4.0 kg with semaglutide and −0.5 kg with placebo in the overall study population (Figure 5A). The absolute BW reduction from baseline to first primary MACE or censoring was numerically larger in participants with higher BMI at baseline than those with lower BMI (Figure 5A), whereas the percentual BW reduction was similar between the BMI subgroups (Figure 5B). The HR for BW change reflects the risk of primary MACE associated with 1 kg change in BW, and results indicate that changes in BW had no effect on the risk of primary MACE in the overall study population or across BMI subgroups (Figure 5C, left forest plot). The HR for treatment indicates the risk of primary MACE observed for semaglutide versus placebo. The treatment effect of semaglutide by baseline BMI subgroups was only slightly affected by incorporating change in BW as a time‐dependent variable in the Cox proportional hazards model (Figure 5C, right forest plot), indicating that the effect of semaglutide on primary MACE risk remained consistent when accounting for various degrees of weight loss.

FIGURE 5.

BW changes from baseline and effect of BW changes or semaglutide treatment on primary MACE risk in the overall study population and by baseline BMI subgroups. The absolute (A) and percentual (B) mean BW change from baseline to first event of primary MACE or censoring was assessed in the overall study population and by baseline BMI subgroups. Censoring was defined as loss to follow‐up, death, or end of study. (C) The association between BW change or treatment effect and the risk of primary MACE was analysed using a Cox proportional hazards model with BW change as a time‐dependent variable, treatment as a fixed factor and baseline BW as covariate (for the overall study population only); the model was stratified by trial (SUSTAIN 6, PIONEER 6) and presence of cardiovascular disease at baseline (yes/no). The HR of BW change reflects the risk of primary MACE associated with a 1 kg change in BW, while the HR of treatment indicates the risk of primary MACE between semaglutide versus placebo. ^†There were 29 participants who were excluded because they had missing data on baseline BW or had no follow‐up BW measurements before first event of primary MACE or censoring. One participant was further excluded in the BMI subgroup analysis due to missing baseline BMI value. %, proportion of participants with event; BMI, body mass index; BW, body weight; CI, confidence interval; HR, hazard ratio; MACE, major adverse cardiovascular events; N, total number of participants; n, number of participants with event.

4. DISCUSSION

This pooled post hoc analysis of the SUSTAIN 6 and PIONEER 6 CVOTs demonstrated that the cardiorenal effects of semaglutide are consistent in people with T2D at high CV risk or with established CVD, regardless of prior CVD and baseline BMI, even when accounting for various degrees of weight loss.

This analysis shows that the risk of primary MACE, expanded MACE, and new or worsening nephropathy was significantly reduced with semaglutide versus placebo in the overall study population. Cardiorenal benefits were consistent across subgroups with and without prior revascularisation, MI, or stroke/TIA. Our findings support previous observations from a CVOT meta‐analysis demonstrating that GLP‐1RAs reduced the overall risk of individual MACE components, all‐cause mortality, hospitalisation for HF, and worsening kidney function versus placebo in T2D ¹⁴ ; moreover, the effect of GLP‐1RAs on MACE risk was consistent regardless of prior established CVD. ¹⁴ Similarly, previous observations showed that both formulations of semaglutide reduced the relative risk of primary MACE versus placebo in people with T2D across varying CV risk and established CVD and/or CKD. ¹⁷ , ²⁶

This study also shows that differences in baseline BMI, even BMI <25 kg/m², do not impact the effect of semaglutide on the evaluated CV and kidney‐related outcomes. Furthermore, in both the overall population and across baseline BMI subgroups, BW change had no significant impact on primary MACE risk, with semaglutide treatment consistently reducing the risk of primary MACE irrespective of changes in BW. This is in line with two large meta‐analyses of GLP‐1RA CVOTs that showed how the benefit of GLP‐1RAs on MACE risk was similar across baseline BMI subgroups in people with T2D at high CV risk or with established CVD. ¹⁴ , ²⁷ Similarly, data from the SELECT CVOT have shown that OW s.c. semaglutide 2.4 mg reduced the risk of primary MACE by 20% (p < 0.001 for superiority) versus placebo, regardless of baseline BMI or different forms and degrees of CVD in a population with overweight or obesity and established CVD, but without diabetes. ²⁸ A post hoc analysis of the LEADER and SUSTAIN 6 CVOTs, separately evaluating the effect of OD s.c. liraglutide and OW s.c. semaglutide, respectively, demonstrated that the risk reduction with OW s.c. semaglutide versus placebo was consistent for expanded MACE, CV death, and a kidney composite endpoint, but was inconsistent for primary MACE between baseline BMI subgroups. ¹⁸ However, this inconsistency was nominally significant, and a complementary analysis confirmed benefits of semaglutide versus placebo across a continuum of BMI values. ¹⁸ As the current pooled post hoc analysis included data for OW s.c. and OD oral semaglutide from the SUSTAIN 6 and PIONEER 6 CVOTs, respectively, our findings on consistent risk reduction of CV outcomes regardless of baseline BMI support previous conclusions.

The precise mechanisms by which GLP‐1RAs exert beneficial CV and kidney‐related effects remain unknown. Direct mechanisms explored ex vivo, in vitro, or in animal models include inhibition of endothelial cell dysfunction, attenuation of abnormal migration, proliferation, and apoptosis of vascular smooth muscle cells, and reduction of macrophage inflammation. ²⁹ Semaglutide treatment might alter the intra‐cardiac expression of inflammation‐related genes, as demonstrated in an atherosclerotic animal model. ³⁰ In addition, GLP‐1RAs might reduce intrarenal sodium reabsorption, hypoxia, oxidative stress, inflammation, and cell death, while increasing natriuresis and haemodynamics, and modifying neural and renin–angiotensin–aldosterone system signalling. ¹³ , ²¹

The SUSTAIN 6, PIONEER 6, and SELECT CVOTs demonstrated that people with T2D or overweight/obesity achieved greater weight loss and improvements in CV risk factors (such as blood pressure and lipids) with semaglutide versus placebo. ¹⁵ , ¹⁶ , ²⁸ Effects of semaglutide on BW might contribute to indirect beneficial effects on CV and kidney‐related outcomes. ¹³ , ²⁰ , ²¹ However, the current analysis shows that the effect of semaglutide on the risk of primary MACE versus placebo in people with T2D and different baseline BMI remained consistent even when accounting for changes in BW, supporting results in the main analysis. This implies that semaglutide might act independently of weight loss. This observation aligns with previous analyses showing that CV and kidney‐related benefits of GLP‐1RAs versus placebo are partly influenced by improved HbA1c, reduced albuminuria, and reduced systolic blood pressure, but with no or minimal influence of BW changes. ²² , ²³ , ²⁴

The key strength of this post hoc analysis is that a relatively large and heterogenous study population with various comorbidities was included. Hence, the results are relevant for clinical practice with high transferability to real‐world settings, noting that semaglutide was efficacious in people with T2D, with different experiences of prior CVD and a wide range of baseline BMIs, including those with baseline BMI <25 kg/m². Furthermore, data for OW s.c. and OD oral semaglutide were pooled, allowing the efficacy of both formulations of semaglutide on CV outcomes to be assessed.

This study provides an overview of cardiorenal effects of OW s.c. and OD oral semaglutide in a heterogeneous T2D population with variable BMI and CVD baseline characteristics, although larger and comprehensive trials are needed to better understand the effect of semaglutide on hard CV or kidney‐related outcomes. Indeed, the FLOW kidney outcomes trial demonstrated that OW s.c. semaglutide 1.0 mg has beneficial effects on kidney outcomes and parameters versus placebo in people with T2D and CKD, as shown by a 24%‐risk reduction of major kidney disease events (a composite of onset of kidney failure (dialysis, transplantation, or an eGFR of <15 mL/min/1.73 m²), ≥50% reduction in eGFR from baseline, or death from kidney‐related or CV causes) and a slower decrease in mean annual eGFR slope (between‐group difference [95% CI]: 1.16 [0.86; 1.47]; p < 0.001). ¹⁹ The FLOW trial also showed a broad and consistent effect of semaglutide on major kidney disease events in people with different baseline BMI or prior CVD status (MI or stoke and chronic HF). ¹⁹ Preliminary results from the SOUL CVOT (NCT03914326) demonstrated that the risk of primary MACE was significantly reduced by 14% with OD oral semaglutide 14 mg versus placebo in people with T2D and CVD and/or CKD. ³¹

wThe limitations of the study include the post hoc nature of the analyses and the lack of kidney‐related outcome data from PIONEER 6. While our study offers insights into the cardiorenal effects of semaglutide in various subgroups of people with T2D, the analyses were not specifically powered for these endpoints. As such, the consistent cardiorenal effects, based on a p _INT >0.05 for all comparisons, may not necessarily signify equivalence of observed effects given the small number of participants in some subgroups. Additionally, the wide 95% CIs may have limited statistical power to detect subgroup differences. Hence, the lack of statistical significance in some findings may be due to power or sample size limitations rather than the absence of a true effect. Accordingly, these results should be considered hypothesis‐generating and warrant further investigation.

5. CONCLUSION

This study confirms that the beneficial cardiorenal effects of semaglutide versus placebo were consistent in people with T2D at high CV risk regardless of prior CVD and baseline BMI, even when accounting for changes in BW, indicating that semaglutide might have direct effects on the CV system and kidneys. By simultaneously analysing several baseline characteristics that contribute to disease heterogeneity and diabetes‐associated comorbidities, this pooled post hoc analysis supports the broad utility and efficacy of OW s.c. and OD oral semaglutide in a diverse T2D population.

AUTHOR CONTRIBUTIONS

All authors were involved in manuscript conceptualisation, development, and critical review and editing. All were involved in data interpretation, and all authors read and approved the final version of the manuscript.

FUNDING INFORMATION

This analysis was funded by Novo Nordisk (Shanghai) Pharma Trading Co., Ltd., Beijing, China. Writing and editorial assistance was provided by AXON Communications.

CONFLICT OF INTEREST STATEMENT

Jingmin Zhou declares no conflicts of interest. Mansoor Husain has received research grants from AstraZeneca, Merck, and Novo Nordisk; consultancy fees for participation in advisory board meetings from AstraZeneca, Boehringer Ingelheim, Eli Lilly, Janssen, Merck, Novo Nordisk, and Roche; speaker fees from AstraZeneca, Boehringer Ingelheim, Janssen, Merck, and Novo Nordisk; and holds two patents relating to glucagon‐like peptides. Yang Li, Wenyan Liu, and Zewei Shen are employees of Novo Nordisk and own stocks in the company. Tina Vilsbøll has served on scientific advisory panels, been part of speaker's bureaus, served as a consultant to and/or received research support from Amgen, AstraZeneca, Boehringer Ingelheim, Eli Lilly, GlaxoSmithKline, Mundipharma, Novo Nordisk, Roche, Sun Pharmaceuticals, and Zealand Pharma. Junbo Ge declares no conflicts of interest.

Supporting information

Data S1. Supporting Information.

Zhou J, Husain M, Li Y, et al. Effect of semaglutide versus placebo on cardiorenal outcomes by prior cardiovascular disease and baseline body mass index: Pooled post hoc analysis of SUSTAIN 6 and PIONEER 6. Diabetes Obes Metab. 2025;27(10):5706‐5715. doi: 10.1111/dom.16621

DATA AVAILABILITY STATEMENT

The datasets used and/or analysed in the post hoc analysis are available from the corresponding author on reasonable request.