FormalPara Key Summary Points

Why carry out the study?

Type 2 diabetes represents a major clinical burden worldwide, including China

Tirzepatide is a novel glucose-dependent insulinotropic polypeptide and glucagon-like peptide-1 receptor agonist that was shown to improve glycemic control and weight loss in Chinese patients with type 2 diabetes in the SURPASS-AP-Combo trial

This subgroup analysis of the SURPASS-AP-Combo trial assessed the potential influence of baseline characteristics on the efficacy of tirzepatide including age, sex, baseline HbA1c, body mass index, body weight, duration of diabetes, and concomitant oral antihyperglycemic medications. Safety analyses included the incidence of gastrointestinal adverse events (AEs) and hypoglycemia

What was learned from the study?

Tirzepatide was associated with reduced blood glucose and body weight in the different subgroups, consistent with previous reports for tirzepatide

These results support the use of tirzepatide in the Chinese population across different patient subgroups

Introduction

In recent years, there has been epidemic growth in the number of individuals with diabetes worldwide, increasing from 422 million in 2014 to 529 million in 2021 [1, 2], primarily attributed to the escalating prevalence of type 2 diabetes (T2D) [3]. Concerningly, the global prevalence of diabetes is forecast to increase further in the coming years, by 25% by 2030 and 51% by 2045 [4]. Current data suggest that 60% of people with diabetes reside in Asia, with the highest prevalence in East Asia [5]. In 2021, the overall incidence of diabetes in China was estimated at > 140 million [6]. This rising prevalence of diabetes represents a major burden to patients, caregivers and healthcare systems worldwide [6]. Notably, the adult population affected by T2D is heterogenous, exhibiting diverse clinical characteristics and associated comorbidities. For example, East-Asian patients with T2D have a younger age of onset and lower body mass index (BMI) ranges compared with white patients [7].

Tirzepatide is a once-weekly, glucose-dependent insulinotropic polypeptide/glucagon-like peptide-1 (GIP/GLP-1) receptor agonist that has demonstrated clinically meaningful improvements in glycemic control and weight loss, along with a favorable safety profile in the global SURPASS 1–6 and Japanese SURPASS J-mono and J-Combo clinical trials, which included patients with T2D across various disease durations and treatment backgrounds [8,9,10,11,12,13,14,15]. Based on the findings from these clinical trials, tirzepatide was the first GIP/GLP-1 receptor agonist approved for treating T2D and is also approved for chronic weight management in the US [16]. Currently, tirzepatide is under investigation for the treatment of T2D and obesity in China.

SURPASS-AP-Combo was, to our knowledge, the first multicenter, phase 3 clinical trial to investigate the efficacy and safety of once-weekly tirzepatide versus insulin glargine in a predominantly Chinese population with T2D [17]. The results from this study were consistent with the findings of the global and Japanese SURPASS trials, demonstrating superior glycemic control and weight reduction with tirzepatide and also a safety profile in line with previous trials of tirzepatide [8,9,10,11,12,13,14,15].

To better understand the effects of once-weekly tirzepatide in an Asia-Pacific, predominately Chinese, population with T2D, we conducted a post hoc subgroup analysis of the SURPASS-AP-Combo to investigate the potential influence of baseline characteristics on the efficacy of once-weekly tirzepatide 5, 10 and 15 mg. Safety analyses included the incidence of gastrointestinal AEs and hypoglycemia.

Methods

Study Design

This was a post hoc analysis of SURPASS-AP-Combo, a randomized, open-label, multicenter, parallel-arm, phase 3 trial (ClinicalTrials.gov registration: NCT04093752). The study design and primary results have been published previously [17]. Briefly, the study recruited patients with T2D from 66 sites across China (n = 43), South Korea (n = 13), Australia (n = 6) and India (n = 4) and comprised a 1-week screening period, a 2-week lead-in period, a 40-week treatment period and a 4-week safety follow-up period. The study protocol was approved by Ethics Review Boards at each site, and written informed consent was collected from all participants. The study was conducted in accordance with the principles of international ethics guidelines, including the Declaration of Helsinki, and applicable local laws and regulations.

Patients and Treatment

The eligibility criteria for this trial have been published previously [17]; a summary of the inclusion and exclusion criteria can be found in the Supplementary Materials. Eligible patients were randomized 1:1:1:1 to receive once-weekly tirzepatide 5, 10, 15 mg or once-daily insulin glargine, stratified by baseline HbA1c (≤ 8.5, > 8.5%), country and use of concomitant oral antidiabetic treatments (metformin alone, metformin plus a sulfonylurea). Tirzepatide was initiated at 2.5 mg for all patients and increased by 2.5 mg every 4 weeks until the target dose was reached. All patients continued background therapy with metformin (± sulfonylurea) at the same pre-study dose. Rescue antihyperglycemic therapy was allowed for patients with severe, persistent hyperglycemia who met the pre-defined criteria.

Study Assessments and Subgroups

The current analyses evaluated efficacy data in patients included in SURPASS-AP-Combo who received tirzepatide (5, 10 and 15 mg). A patient flow chart is provided in Supplementary Fig. 1. Subgroups were defined according to the following key potential influential factors: age (< 65, ≥ 65 years), sex (male, female), baseline HbA1c (≤ 8.5, > 8.5%), baseline BMI (< 25, ≥ 25 kg/m2), baseline body weight (< 75, ≥ 75 kg), duration of diabetes (< 10, ≥ 10 years) and concomitant oral antidiabetic medication (OAM) use (metformin, metformin plus sulfonylurea). Baseline characteristics and the following efficacy endpoints at week 40 were evaluated across the predefined subgroups: change from baseline in HbA1c, body weight, fasting serum glucose (FSG) and daily glucose average calculated from 7-point self-measured blood glucose (SMBG) profiles. Safety analyses included the incidence of gastrointestinal adverse events (AEs) and hypoglycemia.

Statistical Analysis

Efficacy analyses of change from baseline to week 40 were conducted using a mixed model for repeated measures (MMRM) with treatment, subgroup, country, baseline OAM use (metformin, metformin plus sulfonylurea), baseline HbA1c category (≤ 8.5, > 8.5%), subgroup, time, subgroup-by-time interaction, treatment-by-time interaction, treatment-subgroup-time three-way interaction and baseline value as covariates. Least-square (LS) means, 95% confidence intervals (CIs) and P values for the comparisons between subgroups were computed from the MMRM model. For analyses of HbA1c, the baseline HbA1c category was not included in the model. The subgroups by baseline HbA1c (≤ 8.5, > 8.5%) and concomitant OAM use (metformin, metformin plus sulphonylurea) were only included once in the model. An unstructured covariance structure was used to model the relationship of intra-patient errors. Gastrointestinal AEs and hypoglycemia were reported as percentages, calculated based on the number of patients in each subgroup category.

Efficacy was evaluated using data obtained during the treatment period (from all randomized tirzepatide treated patients who received ≥ 1 dose of study drug) and excluding data after initiating rescue antihyperglycemic medication or discontinuation of study drug. Safety analyses used all data collected from the start of treatment to the end of the safety follow-up period. All tests of treatment effect were conducted at a two-sided alpha level of 0.05. All analyses were performed using SAS Version 9.4.

Results

Baseline Characteristics

A total of 687 patients were included for this subgroup analysis, who received tirzepatide 5 mg (n = 230), tirzepatide 10 mg (n = 228) or tirzepatide 15 mg (n = 229). A greater proportion of patients were male versus female across the tirzepatide treatment subgroups (5, 10 and 15 mg) (Table 1). The mean (SD) overall age of patients across the tirzepatide treatment groups ranged from 53.1 (11.2) to 54.3 (11.6) years. The mean (SD) BMI of patients across the tirzepatide treatment groups ranged from 27.7 (3.8) to 28.1 (3.9) kg/m2. Overall, baseline characteristics were generally well balanced across the tirzepatide treatment groups (Table 1).

Table 1 Summary of baseline characteristics
Full size table

Efficacy

Changes from Baseline in HbA1c

All doses of tirzepatide (5, 10 and 15 mg) resulted in a reduction in HbA1c from baseline to week 40 across each subgroup, irrespective of age, sex, baseline HbA1c, BMI, body weight, duration of diabetes and concomitant OAM use (Fig. 1). For the individual analyses by subgroup, LS mean (standard error [SE]) reductions in HbA1c from baseline to 40 weeks were significantly greater in patients with a higher baseline HbA1c (≤ 8.5 vs. > 8.5%) with all tirzepatide doses: tirzepatide 5 mg – 2.52% (0.093) vs. – 1.76% (0.105) P < 0.001; tirzepatide 10 mg – 2.95% (0.099) vs. – 1.78% (0.103), P < 0.001; tirzepatide 15 mg – 3.01% (0.100) vs. – 1.86% (0.102), P < 0.001 (Fig. 1). LS mean (SE) reductions in HbA1c were significantly greater in younger patients (< 65 vs. ≥ 65 years) with tirzepatide 15 mg: – 2.53% (0.087) vs. – 2.08% (0.160), respectively, P = 0.011 (Fig. 1). Furthermore, a numerically greater LS mean (SE) reduction in HbA1c was observed in patients with a higher baseline body weight (≥ 75 vs. < 75 kg) with tirzepatide 10 and 15 mg, and the difference reached statistical significance for tirzepatide 10 mg: – 2.56% (0.111) vs. – 2.26% (0.105), respectively, P = 0.040 (Fig. 1). Comparatively small and non-statistically significant differences in the reduction in HbA1c from baseline to week 40 were observed between the remaining subgroups across the tirzepatide treatment groups (P ≥ 0.05).

Fig. 1
figure 1

Changes from baseline to week 40 in HbA1c (%) stratified by potential influential factors. P-value from MMRM model. BMI body mass index; HbA1c hemoglobin A1c; LS mean least-square mean; MMRM mixed-model for repeated measures; n number of patients who were randomized and received at least one dose of study drug; OAM oral antihyperglycemic medication; TZP tirzepatide

Full size image

Change from Baseline in Body Weight

Across all subgroups, tirzepatide was associated with a reduction in mean body weight from baseline to week 40, irrespective of age, sex, baseline HbA1c, BMI, body weight, duration of diabetes and use of concomitant OAMs (Fig. 2). LS mean (SE) reductions in body weight from baseline to 40 weeks were significantly greater in patients with a higher baseline body weight (≥ 75 vs. < 75 kg) across all tirzepatide doses: tirzepatide 5 mg – 5.83 kg (0.460) vs. – 4.13 kg (0.493), respectively, P = 0.010; tirzepatide 10 mg – 7.89 kg (0.511) vs. – 6.34 kg (0.473), respectively, P = 0.022; tirzepatide 15 mg – 8.09 kg (0.491) vs. – 6.31 kg (0.485), respectively, P = 0.008 (Fig. 2). Similarly, patients with a higher baseline BMI (≥ 25 vs. < 25 kg/m2) receiving tirzepatide 10 and 15 mg also experienced a numerically greater LS mean (SE) reduction in body weight, and this was statistically significant in the 5 mg treatment group: – 5.42 kg (0.385) vs. – 3.66 kg (0.714), P = 0.026 (Fig. 2). A lower baseline HbA1c (≤ 8.5% vs. > 8.5%) was associated with a significantly greater LS mean (SE) reduction in body weight from baseline to week 40 with tirzepatide 5 mg: – 5.85 kg (0.509) vs. – 4.41 kg (0.447), respectively, P = 0.029 (Fig. 2). In addition, LS mean (SE) reductions in body weight were numerically greater in younger patients (< 65 vs. ≥ 65 years) across all tirzepatide treatment groups, and the difference reached significance for patients receiving tirzepatide 15 mg: – 7.90 kg (0.393) vs. – 4.67 kg (0.728), respectively, P < 0.001 (Fig. 2). No statistically significant differences in reduction in body weight were observed in the remaining subgroups across the tirzepatide treatment groups (P ≥ 0.05).

Fig. 2
figure 2

Changes from baseline to week 40 in body weight (kg) stratified by potential influential factors. P-value from MMRM. BMI body mass index; HbA1c hemoglobin A1c; LS mean least-square mean; MMRM mixed-model for repeated measures; n number of patients who were randomized and received at least one dose of study drug; OAM oral antihyperglycemic medication; TZP tirzepatide

Full size image

Change from Baseline in FSG

Tirzepatide (5, 10 and 15 mg) was associated with a reduction in mean FSG from baseline to week 40, irrespective of age, sex, baseline HbA1c, BMI, body weight, duration of diabetes or use of concomitant OAMs (Fig. 3). A greater LS mean (SE) reduction in FSG from baseline to week 40 was observed in patients with a lower baseline HbA1c (≤ 8.5% vs. > 8.5%) receiving tirzepatide 5 mg: – 63.16 mg/dl (3.678) vs. – 53.82 mg/dl (3.376), respectively, P = 0.033 (Fig. 3). Similar reductions in FSG from baseline to week 40 were observed in the remaining subgroups, across the tirzepatide treatment groups (P ≥ 0.05) (Fig. 3).

Fig. 3
figure 3

Changes from baseline to week 40 in FSG (mmol/l) stratified by potential influential factors. P-value from MMRM. BMI body mass index; FSG fasting serum glucose; HbA1c hemoglobin A1c; LS mean least-square mean; MMRM mixed-model for repeated measures; n number of patients who were randomized and received at least one dose of study drug; OAM oral antihyperglycemic medication; TZP tirzepatide

Full size image

Reduction in Daily Glucose Average of SMBG Profiles

At week 40, all doses of tirzepatide were associated with reductions from baseline in the daily glucose average from SMBG profiles, irrespective of age, sex, baseline HbA1c, BMI, body weight, duration of diabetes and use of concomitant OAMs (Supplementary Fig. 2). A significantly greater LS mean (SE) reduction in daily glucose average from SMBG profiles was observed in younger patients (< 65 vs. ≥ 65 years) receiving tirzepatide 15 mg: – 5.04 mmol/l (0.180) vs. –4.18 mmol/l (0.288), respectively, P = 0.003 (Supplementary Fig. 2). Among patients receiving tirzepatide 5 mg, a lower baseline HbA1c (≤ 8.5% vs. > 8.5%) was associated with a significantly greater LS mean (SE) reduction in daily glucose average from SMBG profile: – 4.75 mmol/l (0.214) vs. – 4.15 mmol/l (0.198), respectively, P = 0.014 (Supplementary Fig. 2). Conversely, in the tirzepatide 10 and 15 mg treatment groups, a numerically greater reduction in SMBG profiles was observed in patients with a higher baseline HbA1c, but the difference did not reach statistical significance (Supplementary Fig. 2). In the remaining subgroups, similar reductions in SMBG profiles from baseline to week 40 were observed across the tirzepatide treatment groups (P ≥ 0.05) (Supplementary Fig. 2).

Safety

Gastrointestinal AEs and Hypoglycemia Incidence

The proportion of patients reporting treatment-emergent gastrointestinal AEs was generally similar across each subgroup (Table 2). Across all tirzepatide doses, the total incidence of hypoglycemia was broadly comparable among the patient subgroups (Table 3). Patients with a BMI of < 25 vs. ≥ 25 kg/m2 and a body weight of < 75 vs. ≥ 75 kg experienced a higher incidence of hypoglycemia with tirzepatide 10 mg (Table 3). Notably, across all tirzepatide doses, patients using metformin plus sulfonylurea versus metformin alone experienced a higher incidence of hypoglycemia (Table 3).

Table 2 Incidence of treatment-emergent gastrointestinal AEs at 40 weeks stratified by potential influential factors
Full size table
Table 3 Incidence of hypoglycemia at 40 weeks stratified by potential influential factors
Full size table

Discussion

This post hoc subgroup analysis of SURPASS-AP-Combo is the first to our knowledge to evaluate the efficacy of once-weekly tirzepatide stratified by baseline characteristics in an Asian-Pacific, predominantly Chinese, population. All three doses of tirzepatide were associated with reductions in mean HbA1c, body weight, FSG and daily glucose average from SMBG profiles across all investigated subgroups. These findings are consistent with the primary results of SURPASS-AP Combo [17]. Positive associations were observed between several patient subgroups and mean reduction in HbA1c, body weight, FSG and daily glucose average from SMBG profiles from baseline to week 40, which may highlight potential patient populations that may benefit from tirzepatide treatment.

In this analysis, reductions in mean HbA1c from baseline to week 40 were most influenced by patient HbA1c levels at baseline; patients with a higher baseline HbA1c experienced a significantly greater reduction in HbA1c across all tirzepatide doses. A similar observation was reported in a post hoc analysis of the global phase 3 SURPASS clinical trial program that evaluated glycemic control by baseline HbA1c (≤ 8.5% or > 8.5%) [18]. In the present study, a numerically greater reduction in HbA1c from baseline to week 40 was achieved by patients with a higher baseline body weight receiving tirzepatide 10 and 15 mg, and the difference reached statistical significance in patients receiving tirzepatide 10 mg. This finding may suggest an added value of weight loss on glucose reduction in patients with T2D and overweight or obesity. In this study, a significantly greater reduction in HbA1c and SMBG profile was observed in younger patients receiving tirzepatide 15 mg. Interestingly, a previous study in patients with uncontrolled T2D on insulin showed that the glucose-lowering effect of liraglutide was dependent on beta cell function [19, 20]. However, we observed that blood glucose control, including reductions in HbA1c, FSG and SMBG profiles, was similar across the three tirzepatide treatment groups by duration of T2D (< 10 vs. ≥ 10 years), which is consistent with the results of a previous subgroup analysis from the global phase 3 SURPASS studies that investigated the glycemic effect of tirzepatide by duration of diabetes [21]. Additional analysis is warranted to investigate the correlation between the glucose-lowering effect, patient age and the duration of diabetes as well as beta-cell function.

Our results show that all doses of tirzepatide were associated with improvements in body weight, FSG and daily glucose average from SMBG profiles across the patient subgroups at week 40. Notably, reductions in body weight were significantly greater in patients with a higher baseline body weight across all doses of tirzepatide and in the tirzepatide 5 mg group for patients with a higher baseline BMI. These findings were consistent with a post hoc analysis of SURPASS-1 through -5, which showed that absolute weight change was generally greater among patients with higher BMI categories [22]. In the present study, a statistically significant reduction in body weight was observed in patients with a lower baseline HbA1c in the tirzepatide 5 mg group, consistent with results observed with GLP-1 RAs [23, 24]. In the present analysis, the reduction in body weight from baseline was significantly greater in younger patients (< 65 years) receiving tirzepatide 15 mg. However, this result was not consistently observed across the tirzepatide treatment groups. In the baseline HbA1c subgroup (≤ 8.5% vs. > 8.5%), a statistically significant greater reduction in FSG and SMBG profiles was observed in patients with lower baseline HbA1c with tirzepatide 5 mg.

The overall incidence of treatment-emergent gastrointestinal AEs and hypoglycemia at week 40 were generally similar across the different patient subgroups, with some variability observed in specific subgroups. For example, across all tizepatide treatment groups, a higher incidence of hypoglycemia was observed in patients using metformin plus sulphonylurea compared to metformin alone, which is consistent with previous studies [9, 17]. In the present study, baseline BMI (< 25 kg/m2) and body weight (< 75 kg) were associated with a significantly greater incidence of hypoglycemia with tirzepatide 10 mg. Interestingly, in a previous East Asian focused study (predominantly Japanese patients) that investigated the safety and efficacy of tirzeatide according to age and baseline BMI, the rate of AE-related discontinuations was highest within the BMI < 25 kg/m2 and > 65 years of age subgroup across all tirzepatide doses [25]. In our analysis, among patients receiving tirzepatide 5 mg, the incidence of total hypoglycemia was significantly higher in patients with a baseline HbA1c ≤ 8.5 vs. > 8.5%. This finding is similar to previous observations in patients receiving a GLP-1 receptor agonist [23]. In addition, our study found that older patients (> 65 years) experienced a significantly greater total incidence of gastrointestinal AEs with tirzepatide 10 mg. However, in the previously mentioned East Asian study of tirzepatide stratified by age and BMI subgroups, no patterns between gastrointestinal AEs and age were reported [25].

The limitations of this study include those inherent to a post hoc subgroup analysis, such as the uneven disribution of patients within the age and BMI baseline subgroups, with comparably more patients < 65 years old and with a BMI ≥ 25 kg/m2. In addition, as this was a post hoc analysis, the statistical and clinical significance of nominal P values should be interpretted with caution.

Conclusion

In summary, the results of this post hoc subgroup analysis demonstrated improvements in glycemic control, body weight and daily glucose average from SMBG profiles across all tirzepatide doses (5, 10, and 15 mg) regardless of patient age, sex, baseline HbA1c, BMI, body weight, or duration of diabetes or use of concomitant OAMs in a predominantly Chinese population with T2D.