Intense simplified strategy for newly diagnosed type 2 diabetes in patients with severe hyperglycaemia: multicentre, open label, randomised trial

Abstract

Objective: To evaluate whether the intense simplified strategy, which comprises short term intensive insulin therapy (SIIT) followed by subsequent oral antihyperglycaemic regimens, could improve long term glycaemic outcomes in patients with newly diagnosed type 2 diabetes mellitus and severe hyperglycaemia.

Design: Multicentre, open label, randomised trial.

Setting: 15 hospitals in China between December 2017 and December 2020.

Participants: 412 patients with newly diagnosed type 2 diabetes and significant hyperglycaemia (HbA_1c ≥8.5%).

Interventions: All randomised participants initially received SIIT for 2-3 weeks, followed by linagliptin 5 mg/day, metformin 1000 mg/day, combination linagliptin plus metformin, or lifestyle modification alone (control) for 48 weeks.

Main outcome measures: The primary outcome was the percentage of participants achieving HbA_1c <7.0% at week 48 after SIIT. Secondary outcomes included glycaemic control, β cell function, and variations in insulin sensitivity.

Results: 412 participants were randomised. At baseline, the mean age was 46.8 (standard deviation 11.2) years, mean body mass index was 25.8 (2.9), and mean HbA_1c was 11.0% (1.9%). At week 48, 80% (78/97), 72% (63/88), and 73% (69/95) of patients in the linagliptin plus metformin, linagliptin, and metformin groups, respectively, achieved HbA_1c <7.0%, compared with 60% (56/93) in the control group (P=0.02 overall; P=0.003 for linagliptin plus metformin versus control; P=0.12 for linagliptin versus control; P=0.09 for metformin versus control). Additionally, 70% (68/97), 68% (60/88), and 68% (65/95) of patients in the linagliptin plus metformin, linagliptin, and metformin group, respectively, achieved HbA_1c <6.5% compared with 48% (45/93) in the control group (P=0.005 overall; P=0.005 for linagliptin plus metformin versus control; P=0.01 for linagliptin versus control; P=0.008 for metformin versus control; all were significant after adjustment for multiple comparisons). Thus, compared with the control group, participants in the linagliptin plus metformin group were more likely to achieve HbA_1c <7.0% at week 48 (odds ratio 2.78, 95% confidence interval 1.37 to 5.65; P=0.005). Moreover, the linagliptin plus metformin group showed the most significant improvement in fasting plasma glucose and β cell function indices. All treatments were well tolerated.

Conclusions: The intense simplified strategy using subsequent oral therapies post-SIIT, especially the linagliptin plus metformin combination, sustainably improved glycaemic control and β cell function in patients with newly diagnosed type 2 diabetes mellitus and severe hyperglycaemia. This approach offers a promising direction for decision making in the clinical management of type 2 diabetes mellitus.

Trial registration: ClinicalTrials.gov NCT03194945.

Fig 1

Trial profile

Fig 2

Glycated haemoglobin A_1c (HbA_1c) control during follow-up. Proportions of participants achieving HbA_1c <7.0% or HbA_1c <6.5% at week 48. LIN=linagliptin group; LIN+MET=linagliptin plus metformin group; MET=metformin group

Fig 3

Glycated haemoglobin A_1c (HbA_1c) control during follow-up. Change in HbA_1c over time. LIN=linagliptin group; LIN+MET=linagliptin plus metformin group; MET=metformin group; SIIT=short term intensive insulin therapy. *P<0.0167 compared with control group in log-rank tests

Fig 4

Glycated haemoglobin A_1c (HbA_1c) control during follow-up. Change in HbA_1c at week 48 from baseline. LIN=linagliptin group; LIN+MET=linagliptin plus metformin group; MET=metformin group

Fig 5

Glycated haemoglobin A_1c (HbA_1c) control during follow-up. Summarisation of participants in optimal glycaemic control using Kaplan-Meier curves. LIN=linagliptin group; LIN+MET=linagliptin plus metformin group; MET=metformin group; SIIT=short term intensive insulin therapy. *P<0.0167 compared with control group in log-rank tests

Fig 6

Changes of clinical parameters over time in different treatment groups. Variables are compared with generalised estimating equation models adjusted for age, gender, baseline body mass index, and baseline HbA_1c. P<0.0167 compared with control group for LIN+MET in mean FPG and ISSI-2 and for all treatment groups in HOMA-β. 2HPG=two hour postprandial glucose; FPG=fasting plasma glucose; HOMA-β=homoeostasis model assessment of β cell function; HOMA-IR=homoeostasis model assessment of insulin resistance; ISSI-2=insulin secretion-sensitivity index-2; LIN=linagliptin group; MET=metformin group; LIN+MET=linagliptin plus metformin group; SIIT=short term intensive insulin therapy