This 16-week, open-label, parallel-arm, randomized, metformin-controlled trial was conducted between March 2022 and October 2022 at the Department of Endocrinology, Shandong Provincial Hospital, Shandong, China. The Ethics Committee of Shandong Provincial Hospital approved the trial protocol (No. 2022-046/February 2022), which complied with the Declaration of Helsinki. Written informed consent was obtained from all the participants. The trial was registered in the Chinese Clinical Trial Registry (ChiCTR2200057800).

Key inclusion criteria included T2DM diagnosis (according to the 1999 World Health Organization criteria) (18), 18–65 years of age, body mass index (BMI) of ≥25.0 kg/m2, hemoglobin A1c (HbA1c) of 7.0–10.0%, and treated with lifestyle interventions or in combination with a stable dose of one oral hypoglycemic drug (thiazolidinedione, sulfonylurea, glinide, or a-glucosidase inhibitor) for at least 3 months. Key exclusion criteria included type 1 diabetes, gastrointestinal disorders associated with long-term nausea and vomiting, a history of acute or chronic pancreatitis, or had been treated with any GLP-1 RA or dipeptidyl peptidase-4 inhibitor within the last 3 months. Detailed inclusion and exclusion criteria are presented in Supplementary Figure S1 .

Eligible patients were randomly assigned using an Interactive Web Response System in a 2:1 ratio to receive 0.3 mg PEG-Loxe (Hansoh Pharma) or 1500 mg metformin (Merck). Randomization was stratified according to the following two variables:

Moreover, the clinical trial statistician was blinded to the two groups during data analysis.

PEG-Loxe or metformin was added to the current treatment regimen of each patient: OAD therapy or lifestyle interventions. PEG-Loxe was injected subcutaneously once weekly. This treatment followed a fixed-dose-escalation regimen: an initial dose of 0.1 mg for 4 weeks, followed by 0.2 mg for 4 weeks, then a maintenance dose of 0.3 mg for 8 weeks. Similarly, a metformin maintenance dose of 1500 mg was administered in 500 mg weekly increments from 500 mg to 1500 mg.

In cases where the baseline HbA1c level was <7.5%, patients taking sulfonylurea or glinide were asked to decrease their dose to minimize the risk of hypoglycemia (10). Dose adjustment was performed according to the methods of Elhadd et al. and Kendall et al. (19, 20).

Patient visits were at 4, 8, and 16 weeks, and they included physical examination and data collection. The following information was obtained: demographic data, medical history, vital signs, visceral fat area (VFA), and laboratory test results. Laboratory tests included those for HbA1c, fasting plasma glucose (FPG), C-peptide, lipid profile, and liver function. Additionally, any adverse events (AEs) were recorded. For weight measurement, patients were instructed to remain in the fasting condition, wear light clothing, and take off their shoes. Systolic blood pressure (SBP), diastolic blood pressure (DBP), and pulse rate were measured using a blood pressure monitor (HEM-7312; OMRON, Kyoto, Japan). VFA was assessed using bioelectric impedance analysis (InBody 720, Seoul, South Korea). Fasting blood samples were collected in the morning and analyzed at the Clinical Laboratory of Shandong Provincial Hospital. β-cell function (HOMA-B) and insulin sensitivity (HOMA-S) were estimated using FPG and fasting C-peptide in an updated Homeostasis Model Assessment (HOMA2) obtained from the University of Oxford database (https://www.dtu.ox.ac.uk/homacalculator/). Hypoglycemia was classified as level 1, 2, or 3 based on the definitions of the American Diabetes Association guidelines (16) (Supplement).

Endpoints were collected at week 16, and the primary endpoint was a change in body weight. The secondary endpoints included the proportion of patients with ≥5% and ≥10% weight loss percentages; and changes in BMI, waist circumference (WC), VFA, HbA1c, FPG, C-peptide, HOMA-B, HOMA-S, total cholesterol (TC), triglycerides (TG), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), SBP, and DBP. The safety endpoints included AEs, serious AEs (SAEs), hypoglycemic events, levels of alanine aminotransferase (ALT) and aspartate aminotransferase (AST), and pulse rate.

The planned sample size of 150 patients was randomized 2:1 to receive either PEG-Loxe or metformin. This sample size was expected to provide a power of 80% to detect a difference of ≥2 kg weight loss between the PEG-Loxe and metformin groups, with a standard deviation (SD) of 3.66 (17, 21), α level of 0.05, and 20% withdrawal rate.

The normal distribution of variables was evaluated using the Kolmogorov-Smirnov test and by assessment of residual distribution. Baseline variables were analyzed using independent t test, Mann-Whitney U test, and χ² test.

The efficacy analyses were evaluated using the full analysis set, defined as patients exposed to ≥1 treatment dose and had a baseline assessment. Safety data were assessed using the safety analysis set, defined as patients exposed to ≥1 treatment dose. The primary endpoint (change in body weight) was analyzed using a mixed model for repeated measurements (MMRM), which included group, time, and the corresponding interactions as fixed effects, and baseline weight and sex as covariates. MMRM was used to analyze BMI, WC, VFA, HbA1c, FPG, C-peptide, HOMA-B, HOMA-S, TC, TG, LDL-C, HDL-C, SBP, DBP, ALT, AST, and pulse rate. Categorical variables were evaluated using the χ² test or Fisher exact test. Missing data were imputed using a multiple linear imputation analysis according to the rules of Rubin (22). Sensitivity analyses were performed on the per-protocol set, defined as patients who completed the trial without major protocol violations.

Results are shown as adjusted mean and 95% CI, if not indicated otherwise. A P value < 0.05 (two-tailed) was considered statistically significant. All statistical analyses were performed using Statistical Analysis System (SAS) version 9.4 (RRID: SCR_008567).

PEG-Loxe treatment resulted in significant weight loss compared to metformin treatment during the trial period. After 16 weeks, the primary endpoints (least-square mean (LSM) weight loss) were 7.52 kg (8.37%) and 2.96 kg (3.00%) for the PEG-Loxe and metformin groups, respectively, with a between-group mean difference of 4.55 kg (95% CI: 3.43, 5.67; P < 0.001) ( Figures 2A, B ; Table 2 ). Sensitivity analyses showed similar findings ( Supplementary Table S1 ).

After 16 weeks, the proportions of patients with a weight loss of ≥5% were 61.5% and 25.0% in the PEG-Loxe and metformin groups (P < 0.001), respectively, and those of patients with a weight loss of ≥10% were 26.9% and 5.8% in the PEG-Loxe and metformin groups (P =0.001), respectively ( Figures 2C, D ).

The LSM (95% CI) change in BMI from baseline to week 16 was greater in the PEG-Loxe group [-2.55 (-2.74, -2.37) kg/m²] than in the metformin group [-0.92 (-1.17, -0.67) kg/m²] (P < 0.001). The LSM (95% CI) change in WC from baseline to week 16 was greater in the PEG-Loxe treated patients [-12.26 (-12.58, -11.55) cm] than in metformin-treated patients [-5.67 (-6.60, -4.75) cm] (P < 0.001). After 16 weeks, VFA reduction was significantly higher in the PEG-Loxe group [-26.02 (-27.60, -24.44) cm²] than in the metformin group [-12.39 (-14.45, -10.32) cm²] (P < 0.001) ( Table 2 ).

The LSM changes in HbA1c after week 16 were similar between the two groups (PEG-Loxe, -1.22% [95% CI: -1.38 to -1.06]; metformin, -1.17% [95% CI: -1.39, -0.96]; P =0.69). No significant differences were observed in the mean FPG reduction relative to the baseline between the groups at week 16 (PEG-Loxe, -1.46 mmol/L [95% CI: -1.57, -1.34]; metformin, -1.49 mmol/L [95% CI: -1.65, -1.34]; P =0.70). HOMA2-B levels increased in both groups; the change in the PEG-Loxe group was greater than that in the metformin group (P =0.003). In addition, HOMA2-S increased in both groups, with greater changes observed in the metformin group than the PEG-Loxe group (P < 0.001) ( Table 2 ).

TC, TG, and LDL-C levels were improved with PEG-Loxe treatment compared to metformin treatment (P < 0.001). SBP was reduced by 3.18 mmHg and and 0.28 mmHg with PEG-Loxe and metformin, respectively (P < 0.001). DBP was reduced by 1.34 mmHg and 0.19 mmHg with PEG-Loxe and metformin, respectively (P < 0.001) ( Table 2 ). Supplementary Table S2 also shows these efficacy variables at weeks 4 and 8.

After 16 weeks of treatment, the incidence of AEs was 46.2% (48/104) and 44.2% (23/52) in the PEG-Loxe and metformin groups, respectively (P =0.82). The incidence of SAEs was 2.9% (3/104) and 1.9% (1/52) in the PEG-Loxe and metformin groups, respectively. No deaths occurred during the study period. The most common AEs during the 16-week treatment were gastrointestinal disorders, with a greater incidence in the PEG-Loxe group (24.0%) than in the metformin group (17.3%) ( Table 3 ). Gastrointestinal disorders were mostly mild to moderate and occurred primarily during the first four weeks of treatment. In the PEG-Loxe group, the incidence of gastrointestinal disorders was 11.5%, 5.8%, and 6.7% with 0.1 mg, 0.2 mg, and 0.3 mg treatment, respectively. Acute pancreatitis was not reported during this trial. In addition, the incidence of hypoglycemic events was 2.9% (3/104) and 3.8% (2/52) in the PEG-Loxe and metformin groups, respectively. No level 3 hypoglycemia was reported. Eight patients, five (4.8%) in the PEG-Loxe group and three (5.8%) in the metformin group, discontinued treatment because of AEs.

At week 16, the change in ALT was -6.88 U/L (95% CI: -7.54, -6.22) with PEG-Loxe and -2.33 U/L (95% CI: -3.17, -1.48) with metformin (P < 0.001). The changes in AST were -6.01 U/L (95% CI: -6.59, -5.43) with PEG-Loxe and -1.74 U/L (95% CI: -2.51, -0.97) with metformin (P < 0.001). Slight increases in pulse rate were observed in both groups: 2.07 bpm in the PEG-Loxe group and 0.43 bpm in the metformin group (P < 0.001) ( Table 2 ).