De-Intensification from Basal-Bolus Insulin Therapy to Liraglutide in Type 2 Diabetes: Predictive Value of Mean Glycaemia during Fasting Test

Successful conversion from insulin therapy to glucagon-like peptide 1 receptor agonist (GLP-1RA) with basal insulin in well-controlled patients has already been demonstrated. However, the data concerning individuals with poor glycaemic control are scarce. The aim of this work was to assess the success rate of insulin therapy to liraglutide transition in poorly controlled diabetes in a real-world clinical setting and to define predictors of success. We are the first to present the method of a fasting test as a way to identify the patients at higher risk of failure after treatment de-intensification. The retrospective observational study analyzed data of 62 poorly controlled obese diabetic patients on high-dose insulin therapy, who were subjected to a 72 h fasting test during hospitalization and subsequently switched to liraglutide ± basal insulin therapy. During the fasting, all antidiabetic treatment was discontinued. Patients were classified as responders if they remained on GLP-1RA treatment after 12 months. Non-responders restarted the basal-bolus insulin (BBI) regimen. Development of glycated hemoglobin (HbA1c) and body weight in both groups, alongside with parameters associated with the higher risk of return to the BBI regimen, were analyzed. A total of 71% of patients were switched successfully (=responders). Responders had more significant improvement in HbA1c (-6.4 ± 19.7 vs. -3.4 ± 22.9 mmol/mol) and weight loss (-4.6 ± 7.1 vs. -2.5 ± 4.0). Statistically significant difference between groups was found in initial HbA1c (75.6 ± 17.9 vs. 90.5 ± 23.6; = 0.04), total daily dose of insulin (67.6 ± 36.4 vs. 90.8 ± 32.4; = 0.02), and mean glycaemia during the fasting test (6.9 ± 1.7 vs. 8.6 ± 2.2 mmol/L; < 0.01). This study confirms that therapy de-intensification in poorly controlled patients with a BBI regimen is possible. Higher baseline HbA1c, total daily insulin dose, and mean glucose during fasting test are negative predictive factors of successful therapy de-escalation.