Real-world treatment-transition pathways and dosing patterns of insulin degludec/liraglutide in Chinese adults with type 2 diabetes.

Background and aims: To describe real-world treatment-transition pathways, early IDegLira dosing patterns, treatment simplification, and observed short-term clinical outcomes among Chinese adults with type 2 diabetes (T2D). Methods: This multicenter, non-interventional, retrospective observational registry included 1499 adults with T2D who initiated IDegLira in routine practice across 13 provinces and regions in China from April 2022 to December 2023. The five mutually exclusive prior-treatment pathways were non-injectable therapy, basal insulin, multiple daily injections (MDI), insulin plus GLP-1RA, and GLP-1RA-based therapy. Treatment decisions were physician-directed. Eligible patients had documented continuous use of once-daily IDegLira (the fixed-ratio combination [FRC]) for at least 3 months. Early dosing was defined as the starting dose, dose titration, and dose documented at approximately 12 weeks. Exploratory linear models estimated crude and covariate-adjusted between-pathway differences in paired HbA1c and body-weight change. Results: Baseline HbA1c was available for 1374 patients (9.05 ± 1.78%) and week-12 HbA1c for 498 patients (7.44 ± 2.29%). Among 480 patients with paired baseline and week-12 HbA1c data (all enrolled in the final study period), mean HbA1c decreased from 9.11 ± 2.09% to 7.45 ± 2.28% (mean change -1.66 ± 2.23 percentage points; P < 0.001). Among the week-12 evaluable population, 21.08%, 47.39%, 81.73%, and 89.36% had HbA1c ≤ 6.5%, <7.0%, <7.5%, and <8.0%, respectively. Among 487 patients with paired baseline and week-12 body-weight data (all from the final study period), mean body weight decreased from 75.48 ± 11.12 to 72.72 ± 10.56 kg (mean change -2.76 ± 4.93 kg). In exploratory adjusted models, HbA1c reductions were smaller after basal insulin (mean difference +0.82 percentage points, 95% CI 0.37 to 1.27) and MDI (+1.25, 95% CI 0.53 to 1.98) than after non-injectable therapy, whereas no overall adjusted pathway association was observed for body-weight change (global P = 0.606). The mean starting dose was 14.46 ± 4.06 U and the mean week-12/current dose was 16.72 ± 4.55 U among available cases. Prandial insulin discontinuation was recorded in 269/301 MDI patients (89.37%); concurrent paired HbA1c and systematic post-initiation safety data were unavailable for this subgroup. Conclusions: This study describes real-world treatment-transition and dosing patterns for IDegLira across heterogeneous prior therapies. Glycemic findings are based on the week-12 evaluable population (498 of 1499 patients, all enrolled in the final study period); among 480 patients with paired baseline and week-12 HbA1c data, mean HbA1c decreased by 1.66 percentage points. The study had no comparator group and no basis for causal inference. The findings are descriptive and hypothesis-generating, and may inform clinical discussions and future comparative studies.