Use of Liraglutide in the Real World and Impact at 36 Months on Metabolic Control, Weight, Lipid Profile, Blood Pressure, Heart Rate, and Renal Function

PURPOSE: An observational retrospective study was conducted by 2 diabetes clinics in Italy to assess patterns of use and long-term effectiveness of liraglutide on established and emerging parameters. METHODS: Data from 261 patients with type 2 diabetes who started treatment with liraglutide between 2010 and 2014 were collected. Hierarchical linear regression models were applied to assess trends over time of clinical parameters. Factors associated with higher likelihood of dropout were identified through multivariate logistic analysis. FINDINGS: Liraglutide was initiated as a switch in 42.5% of patients and as an add-on in 49.8%; in 7.7% of the patients initiation of liraglutide was associated with a reduction in the number of pharmacologic agents. A statistically significant reduction after 36 months was found for the following parameters (mean change [95% CIs]): glycosylated hemoglobin (HbA1c; -1.01% [1.34% to -0.68%]), fasting blood glucose (-27.5 [-40.6 to -14.4] mg/dL), weight (-2.9 [-4.5 to -1.3] kg), body mass index (-1.13 [-1.76 to -0.50] kg/m2), waist circumference (-1.74 [-3.85 to -0.37] cm), and LDL-C (-24.7 [-36.67 to -12.8] mg/dL). Improvements in systolic (-3.5 mm Hg) and diastolic (-2.3 mm Hg) blood pressures were observed at 24 months. Albuminuria was reduced by -16.6 mg/L during 36 months, although statistical significance was not reached. Glomerular filtration rate and heart rate were unchanged. Reductions in HbA1c between -0.6% and -1.3% were obtained in specific subgroups. Treatment was effective also in patients with >20 years of diabetes duration, although the likelihood of dropout was 6% higher for each additional year of disease duration (RR = 1.06; 95% CI, 1.01-1.12). The likelihood of dropout was almost four times higher for subjects treated with insulin (RR = 3.82; 95% CI, 1.22-11.96) and more than twice for those treated with sulfonylureas (RR = 2.39; 95% CI, 1.16-4.94) compared with patients not treated with these agents. IMPLICATIONS: Liraglutide used in routine clinical conditions maintains its effectiveness on metabolic control and weight after 3 years. Improvements in terms of metabolic control were found when liraglutide was used as both switch and add-on treatment. In addition, improvements were sustained when liraglutide replaced sulfonylureas or insulin. Diabetes duration had no impact on drug efficacy. Long-term benefits relative to blood pressure and LDL-C were also found, which could not be entirely explained by antihypertensive/lipid-lowering treatment intensification. No major effect on renal parameters was documented. Diabetes duration and some concomitant treatments were associated with a higher likelihood of liraglutide discontinuation. These data can contribute to improve appropriateness and cost-effectiveness profile of liraglutide.