FormalPara Key Summary Points

Treatment guidelines for type 2 diabetes (T2D) recommend a comprehensive, individualised approach with multiple therapeutic goals (including HbA1c reduction and body weight loss), as many people with T2D are associated with several comorbid conditions.

Composite endpoints provide physicians with a holistic view of the clinical benefit of treatments, and a foundation for clinical decision-making.

This post hoc analysis of the PIONEER 1–5 and 8 trials evaluated composite endpoints of ≥ 1% reduction in HbA1c and plus either a ≥ 5% or ≥ 10% reduction in body weight with orally administered semaglutide versus comparators in people with T2D.

The odds of achieving clinically relevant reductions in both HbA1c and body weight were significantly greater with orally administered semaglutide versus comparators.

Digital Features

This article is published with digital features, including a graphical abstract, to facilitate understanding of the article. To view digital features for this article, go to https://doi.org/10.6084/m9.figshare.22645201.

Introduction

Glucagon-like peptide 1 receptor agonists (GLP-1RAs) have been shown to reduce glycated haemoglobin (HbA1c) and body weight in people with type 2 diabetes (T2D), both of which are important therapeutic goals [1, 2]. The glucose and body weight lowering benefits and cardiovascular (CV) benefits of GLP-1RAs are becoming increasingly recognised, and GLP-1RAs are included in international T2D treatment recommendations and guidelines across the disease continuum [3]. Semaglutide is a long-acting GLP-1RA available as a once-weekly subcutaneous injection and a once-daily oral tablet for the treatment of people with inadequately controlled T2D [4]. The tablet formulation of semaglutide may remove the administrative burden on people and physicians (as compared to injectables) and facilitate initiation of GLP-1RAs earlier in the course of T2D [4, 5].

Treatment guidelines for T2D recommend a comprehensive, individualised approach with multiple therapeutic goals (including HbA1c reduction and body weight loss), as many people with T2D also have overweight or obesity, which is associated with several comorbid conditions [1, 3]. Composite endpoints are useful to differentiate between treatment options in terms of overall efficacy and safety, thereby providing physicians with a holistic view of the clinical benefit of different treatments [6, 7]. Additionally, composite endpoints can provide a foundation for clinical decision-making, addressing the limitation of basing clinical decisions on a single outcome, such as change in HbA1c [8].

HbA1c reductions of ≥ 1% and body weight loss of ≥ 5% in people with T2D are considered important indicators of a clinically meaningful response to treatment, and have been shown to reduce the risk of diabetes-related complications [1, 9, 10]. Furthermore, the incidence rate of diabetes-related complications has been reported to decrease with every 1% reduction in HbA1c, while a reduction in body weight of ≥ 5% can confer metabolic improvement and reduce the risk of cardiometabolic disease [3, 9, 11]. In this post hoc analysis, data from several trials included in the PIONEER programme were evaluated to determine to what extent people with T2D treated with orally administered semaglutide versus comparators achieved composite endpoints of ≥ 1% reduction in HbA1c plus either a ≥ 5% or ≥ 10% reduction in body weight, as well as separate endpoints of ≥ 1% decrease in HbA1c and a ≥ 5% or ≥ 10% decrease in body weight.

Methods

Trial Designs and Patient Populations

Data from the PIONEER 1–5 and 8 trials were assessed in this post hoc analysis. PIONEER 6 and 7 were excluded due to the use of different trial designs. PIONEER 6 was a CV outcomes trial in a population with established CV disease or at high risk of CV events, while PIONEER 7 included a flexible dose adjustment approach where not all people received orally administered semaglutide 14 mg [12, 13].

Designs and patient populations of the PIONEER 1–5 and 8 trials have been reported previously [14,15,16,17,18,19]. In brief, the PIONEER 1–5 and 8 trials were conducted across 26–78 weeks, and assessed the efficacy and safety of orally administered semaglutide (3 mg, 7 mg or 14 mg) versus placebo or active comparators (a sodium-glucose co-transporter 2 inhibitor [empagliflozin 25 mg], a dipeptidyl peptidase 4 inhibitor [sitagliptin 100 mg] or a GLP-1RA [subcutaneously administered liraglutide 1.8 mg]) in adults with T2D across a variety of background regimens (diet and exercise alone, metformin, 1–2 oral antidiabetic drugs, basal insulin ± metformin or insulin ± metformin) [14,15,16,17,18,19]. The protocols for the PIONEER 1–5 and 8 trials were approved by local independent ethics committees/institutional review boards at each trial site and conform to the provisions of the Declaration of Helsinki; all people provided written informed consent to participate in the trials.

Assessments

All people who received orally administered semaglutide 14 mg or comparators (active or placebo) during the PIONEER 1–5 and 8 trials were included in this analysis. The following individual and composite endpoints were assessed at the end of treatment (EOT; week 26, 52 or 78): achievement of (1) an HbA1c reduction of ≥ 1%; (2) a body weight loss of ≥ 5%; (3) a body weight loss of ≥ 10%; (4) an HbA1c reduction of ≥ 1% and body weight loss of ≥ 5%; and (5) an HbA1c reduction of ≥ 1% and body weight loss of ≥ 10%. In addition, the proportion of people achieving an HbA1c reduction of ≥ 1% and body weight loss of ≥ 5% at week 26 in all trials was assessed to determine whether the results differed from the EOT.

Statistical Analysis

Data for the trial product estimand (on trial product without rescue medication) were analysed. Missing data were accounted for using an analysis of covariance-based sequential multiple imputation model. Separate logistic regression analyses were performed for complete data sets of each study. Endpoints were analysed with treatment, strata (background medication for PIONEER 3–5 and 8, renal function for PIONEER 5 and insulin regimen for PIONEER 8), region and interaction between strata (PIONEER 5 and 8) as categorical fixed effects, and continuous baseline value(s) as covariates. Results were combined by use of Rubin’s rule to draw inference [20].

Results

Baseline Characteristics

Overall, 3506 people with T2D in the PIONEER 1–5 and 8 trials were included. Baseline demographics and clinical characteristics grouped by trial and treatment arm are presented in Supplementary Table S1. Mean duration of diabetes was lowest in PIONEER 1 and highest in PIONEER 8. Likewise, the mean patient age was lowest in PIONEER 1 and highest in PIONEER 5 and 8. Baseline HbA1c and body weight were similar across trials.

Changes from Baseline in HbA1c and Body Weight

Individual Endpoints

In the PIONEER 1–5 and 8 trials, the odds of achieving the HbA1c reduction of ≥ 1% at EOT were significantly greater with orally administered semaglutide 14 mg compared with placebo (PIONEER 1, 4, 5 and 8, all p < 0.0001) and the active comparators empagliflozin 25 mg (PIONEER 2, p < 0.0001), sitagliptin 100 mg (PIONEER 3, p < 0.0001) and liraglutide 1.8 mg (PIONEER 4, p < 0.05) (Fig. 1A).

Fig. 1
figure 1

Proportion of people in PIONEER 1–5 and 8 trials achieving an HbA1c reduction of ≥ 1% (A), achieving a body weight loss of ≥ 5% (B), or achieving a body weight loss of ≥ 10% (C) at EOT. Data are observed proportions for the trial product estimand (on trial product without rescue medication). In panels A and B and for PIONEER 1–4 in panel C, p values are for the EORs for orally administered semaglutide 14 mg versus placebo or the active comparator. For PIONEER 5 and 8 in panel C, EORs could not be calculated as a result of 0 events in the comparator arms; p values for the risk difference were calculated instead. *p < 0.01 versus placebo; **p < 0.001 versus placebo; ***p < 0.0001 versus placebo; †p < 0.05 versus the active comparator; ††p < 0.001 versus the active comparator; †††p < 0.0001 versus the active comparator. EOR estimated odds ratio, EOT end of treatment, HbA1c glycated haemoglobin, met metformin, SGLT2i sodium-glucose co-transporter 2 inhibitor, SU sulphonylurea

Full size image

Similarly, the odds of achieving a body weight loss of ≥ 5% at EOT were significantly greater with orally administered semaglutide 14 mg than with placebo (PIONEER 1, 4, 5 and 8, all p < 0.0001), sitagliptin 100 mg (PIONEER 3, p < 0.0001) and liraglutide 1.8 mg (PIONEER 4, p < 0.0001), while the odds were comparable for orally administered semaglutide 14 mg and empagliflozin 25 mg in PIONEER 2 (Fig. 1B).

The odds of achieving a body weight loss of ≥ 10% at EOT were significantly greater with orally administered semaglutide 14 mg than with placebo (PIONEER 1, p < 0.01; PIONEER 4, p < 0.001; PIONEER 5, p < 0.001; PIONEER 8, p < 0.01), empagliflozin 25 mg (PIONEER 2, p < 0.001), sitagliptin 100 mg (PIONEER 3, p < 0.0001) and liraglutide 1.8 mg (PIONEER 4, p < 0.001) (Fig. 1C).

Composite Endpoints

The odds of achieving the composite endpoint of an HbA1c reduction of ≥ 1% and body weight loss of ≥ 5% were significantly greater with orally administered semaglutide 14 mg than with any of the comparators in PIONEER 1–5 and 8, including empagliflozin in PIONEER 2, at week 26 and at EOT (all p < 0.0001; Fig. 2A, B).

Fig. 2
figure 2

Proportion of people achieving an HbA1c reduction of ≥ 1% and body weight loss of ≥ 5% at EOT (A), or achieving an HbA1c reduction of ≥ 1% and body weight loss of ≥ 5% at week 26 (B). Data are observed proportions for the trial product estimand (on trial product without rescue medication). *p < 0.0001 for the EOR with orally administered semaglutide 14 mg versus placebo; †p < 0.0001 for the EOR with orally administered semaglutide 14 mg versus the active comparator. EOR estimated odds ratio, EOT end of treatment, HbA1c glycated haemoglobin, met metformin, SGLT2i sodium-glucose co-transporter 2 inhibitor, SU sulphonylurea

Full size image

The odds of achieving both an HbA1c reduction of ≥ 1% and body weight loss of ≥ 10% at EOT were significantly greater with orally administered semaglutide 14 mg than with placebo (PIONEER 1, 4 and 5, all p < 0.01; PIONEER 8, p < 0.05), empagliflozin 25 mg (PIONEER 2, p < 0.0001), sitagliptin 100 mg (PIONEER 3, p < 0.0001) and liraglutide 1.8 mg (PIONEER 4, p < 0.001) (Fig. 3).

Fig. 3
figure 3

Proportion of people achieving an HbA1c reduction of ≥ 1% and body weight loss of ≥ 10% at EOT. Data are observed proportions for the trial product estimand (on trial product without rescue medication). For PIONEER 1–4, p values are for the EORs for orally administered semaglutide 14 mg versus placebo or the active comparator. For PIONEER 5 and 8, EORs could not be calculated as a result of 0 events in the comparator arms; p values for the risk difference were calculated instead. *p < 0.05 versus placebo; **p < 0.01 versus placebo; ***p < 0.001 versus placebo; †p < 0.001 versus the active comparator; ††p < 0.0001 versus the active comparator. EOR estimated odds ratio, EOT end of treatment, HbA1c glycated haemoglobin, met metformin, SGLT2i sodium-glucose co-transporter 2 inhibitor, SU sulphonylurea

Full size image

Discussion

This post hoc analysis evaluated the composite endpoint of a body weight loss of ≥ 5% or ≥ 10% together with an HbA1c reduction of ≥ 1%, arguably a more clinically relevant endpoint than that reported in the individual trials [14,15,16,17,18,19]. Indeed, current T2D treatment guidelines recommend multiple therapeutic goals, including HbA1c < 7% with low or no incidence of hypoglycaemia and ≥ 5% weight loss in people with T2D with overweight or obesity [3, 10, 21]. Composite endpoints are therefore preferred over individual endpoints for their ability to assess the net clinical benefit of an intervention by combining two or more events in one outcome. The extent of information achieved on the efficacy and safety of an intervention by analysing the composite endpoints may prove helpful to physicians and patients as a foundation for clinical decision-making [6].

Orally administered semaglutide 14 mg was better than placebo and active comparators in achieving a combined HbA1c reduction of ≥ 1% and body weight loss of ≥ 5% at week 26 and EOT. Orally administered semaglutide was also better than placebo and active comparators when a larger body weight loss target of ≥ 10% was applied. The odds of achieving each target were also statistically significantly greater with orally administered semaglutide than with comparator.

Studies have previously shown that HbA1c reductions of ≥ 1% and body weight loss of ≥ 5%, measured separately as individual endpoints, are associated with clinically meaningful health benefits [1, 2, 22]. Indeed, the UK Prospective Diabetes Study (UKPDS) of patients with T2D reported that each 1% reduction in HbA1c has the potential to reduce the risk of microvascular complications by 37%, myocardial infarction by 14% and diabetes-related deaths by 21% [9]. Reductions of body weight of ≥ 5% improve glycaemic control, lipid levels and blood pressure in people with overweight/obesity and T2D [1], and are recommended in the American Diabetes Association (ADA) guidelines. The guidelines also recommend considering the effect of medications on weight when selecting a glucose-lowering medication [3]. Furthermore, HbA1c and body weight are considered surrogate endpoints and powerful predictors of CV disease and chronic kidney disease in T2D management. A meta-analysis of five randomised controlled trials (ACCORD, ADVANCE, PROactive, UKPDS and VADT) showed that a reduction in HbA1c of approximately 1% can lead to a 17% relative risk reduction in non-fatal myocardial infarction, while during the UKPDS study, a 33% risk reduction in incident microalbuminuria was reported for patients who achieved glycaemic control [23, 24]. Similarly, weight loss > 5% has been associated with improvements in lipid levels and blood pressure [3, 11, 25]. Evaluating these surrogate endpoints can help physicians draw clinically important conclusions about a therapeutic intervention because of their proven association with these outcomes. The composite endpoint of HbA1c reduction of ≥ 1% and body weight loss of ≥ 5% could therefore be used in clinical practice to differentiate between T2D treatment options by comparing overall efficacy, as part of a holistic treatment approach. Furthermore, orally administered semaglutide provides an option for people who prefer an oral medication over an injectable agent.

A key limitation of this study is that the individual PIONEER trials were not powered for this analysis; therefore, further investigation is warranted to determine whether this composite endpoint can be used as a basis for clinical decision-making in a real-world setting.

Conclusion

Significantly more people achieved clinically relevant reductions in both HbA1c and body weight with orally administered semaglutide versus comparators in the PIONEER 1–5 and 8 trials.