Basal-bolus insulin (BBI) is the most established treatment strategy for patients with very elevated HbA1c (>10%), especially if symptomatic. The combination of metformin (Met), GLP-1 receptor agonist (GLP-1RA) and basal insulin (BI) is a very effective glucose lowering strategy, but it has not been evaluated in patients with very elevated HbA1c. This is the first randomized trial comparing a GLP-1RA+BI treatment regimen to a BBI treatment regimen in patients with very uncontrolled (HbA1c>10%) T2D. Basal insulin detemir was either continued at the prior dose or initiated at 0.3 units/kg and self-titrated. GLP-1RA liraglutide was titrated according to label to 1.8 mg/day. Meal-time insulin aspart was initiated at 0.1 units/kg/meal and self-titrated. Metformin was continued or (re)initiated if tolerated, all other hypoglycemic agents were discontinued. The primary outcome was change in HbA1c at 6 months, analyzed using a mixed model repeated measures analysis. We randomized 120 patients with a mean age (SD) 47.4 (9.5) years, Hispanic 48%, diabetes duration of 10.54 (7.19) years, 76% were already treated with insulin, HbA1c 12.1 (1.4) %, BMI 37.2 (10.3) kg/m2, 86% had one or more symptoms of hyperglycemia.
Both treatment strategies lead to dramatic improvements in HbA1c, treatment with GLP-1RA+BI, compared with BBI, resulted in significantly better glycemic control, weight, insulin dosage, and hypoglycemia (Table).
| Table. Effects of treatment with GLP-1RA + basal insulin (GLP1RA+BI) compared with basal-bolus insulin (BBI) on primary & secondary outcomes. | ||||||
| Variable | GLP1RA+BI Group | BBI Group | Estimated Treatment Difference | P value between groups | ||
| Baseline | 6 months | Baseline | 6 months | |||
| HbA1c(%) | 12.2 (11.8 to 12.6) | 8.1 (7.4 to 8.7) | 11.8 (11.5 to 12.2) | 8.8 (6.0 to 9.1) | 1.1 (0.1 to 2.0) | 0.03 |
| Weight (kg) | 98.2 (89.7 to 106.6) | 97.5 (89.1 to 106.0) | 104.5 (96.2 to 112.8) | 107.6 (99.3 to 115.9) | 3.7 (1.53 to 5.9) | <0.01 |
| TDD Insulin (U/day) (median, 25th to 75th percentile) | 25.0 (22.0 to 36.0) | 64.0 (39.5 to 80.0) | 60.0 (43.0 to 72.0) | 100.0 (66.0 to 155.0) | 0.3* (0.1 to 0.5) | 0.01 |
| HbA1c ≤7% N (%) | NA | 20 (44.4) | NA | 9 (20.5) | Odds Ratio 0.32 (0.13 to 0.82) | 0.02 |
| Patients with any hypoglycemia (glucose <70 mg/dL) N (%) | NA | 19 (35.2) | NA | 37 (66.1) | NA | <0.01 |
| Data represents Least Squares Means (95% confidence intervals) unless otherwise noted. | ||||||
| Table. Effects of treatment with GLP-1RA + basal insulin (GLP1RA+BI) compared with basal-bolus insulin (BBI) on primary & secondary outcomes. | ||||||
| Variable | GLP1RA+BI Group | BBI Group | Estimated Treatment Difference | P value between groups | ||
| Baseline | 6 months | Baseline | 6 months | |||
| HbA1c(%) | 12.2 (11.8 to 12.6) | 8.1 (7.4 to 8.7) | 11.8 (11.5 to 12.2) | 8.8 (6.0 to 9.1) | 1.1 (0.1 to 2.0) | 0.03 |
| Weight (kg) | 98.2 (89.7 to 106.6) | 97.5 (89.1 to 106.0) | 104.5 (96.2 to 112.8) | 107.6 (99.3 to 115.9) | 3.7 (1.53 to 5.9) | <0.01 |
| TDD Insulin (U/day) (median, 25th to 75th percentile) | 25.0 (22.0 to 36.0) | 64.0 (39.5 to 80.0) | 60.0 (43.0 to 72.0) | 100.0 (66.0 to 155.0) | 0.3* (0.1 to 0.5) | 0.01 |
| HbA1c ≤7% N (%) | NA | 20 (44.4) | NA | 9 (20.5) | Odds Ratio 0.32 (0.13 to 0.82) | 0.02 |
| Patients with any hypoglycemia (glucose <70 mg/dL) N (%) | NA | 19 (35.2) | NA | 37 (66.1) | NA | <0.01 |
| Data represents Least Squares Means (95% confidence intervals) unless otherwise noted. | ||||||
*ETD for the log transformed value
M. Abreu: None. A. Tumyan: None. A. Elhassan: None. O. Papacostea: None. K. Peicher: None. P. Dimachkie: None. M.S. Siddiqui: None. B. Adams-Huet: None. X. Li: None. L. Pop: None. I. Lingvay: Advisory Panel; Self; Novo Nordisk A/S, Eli Lilly and Company, AstraZeneca, Intarcia Therapeutics, Inc., Sanofi-Aventis. Research Support; Self; Novo Nordisk A/S, Merck Sharp & Dohme Corp., Pfizer Inc., GI Dynamics Inc., Novartis AG. Other Relationship; Self; Sanofi, AstraZeneca, Boehringer Ingelheim GmbH, Novo Nordisk A/S.
