Effects of liraglutide vs. lifestyle changes on soluble suppression of tumorigenesis-2 (sST2) and galectin-3 in obese subjects with prediabetes or type 2 diabetes after comparable weight loss

Abstract

Background: Soluble suppression of tumorigenesis-2 (sST2) and galectin (Gal)-3 are two biomarkers related to inflammation, metabolic disturbances and to myocardial fibrosis that characterize several cardiac pathological conditions. Increased circulating levels of these molecules have been associated with risk of cardiovascular death. Treatment with liraglutide, a glucagon-like peptide 1 analog, is associated with weight loss, improved glycemic control, and reduced cardiovascular risk. We wanted to assess (I) potential differences between subjects with prediabetes or type 2 diabetes mellitus (T2DM) and healthy controls in sST2 and Gal-3 circulating levels, and their relationship with glycemic control and markers of beta cell function and myocardial injury; (II) whether liraglutide treatment modulates these markers in subjects with prediabetes or early T2DM independently of weight loss; (III) whether baseline levels of any of these two molecules may predict the response to liraglutide treatment.

Methods: Forty metformin-treated obese subjects (BMI ≥ 30) with prediabetes [impaired fasting glucose (IFG) or impaired glucose tolerance (IGT) or both (n = 23)] or newly diagnosed T2DM (n = 17), were randomized to liraglutide or lifestyle counseling until achieving a comparable weight loss (7% of initial body weight). Thirteen subjects were enrolled as healthy controls for baseline sST2 and Gal-3 levels.

Results: Baseline sST2 levels were comparable between controls and obese patients (p = 0.79) whereas Gal-3 levels were significantly higher in patients as compared to controls (p < 0.001). Liraglutide treatment, but not weight loss achieved by lifestyle counseling, decreased plasma sST2 levels (- 9%, beta = - 14.9, standard deviation 6.9, p = 0.037) while Gal-3 levels did not change. A reduction in serum hs-Troponin I was observed after intervention, due to a 19% (p = 0.29) increase in the lifestyle arm, and a 25% decrease (p = 0.033) in the liraglutide arm (between-group difference p = 0.083). Lower baseline Gal-3 levels predicted a better improvement in beta cell function after liraglutide treatment.

Conclusions: Liraglutide-induced reduction in sST2 and possibly hs-TnI suggests that in obese patients with prediabetes or early T2DM this drug may have a positive effect on (cardiac) fibrosis, whereas plasma level of Gal-3 before liraglutide initiation may predict response to the drug in terms of beta cell function improvement. Trial registration Eudract: 2013-001356-36.

Keywords: Cardiac fibrosis; Diabetes; Gal-3; Liraglutide; Markers; sST2.

Fig. 1

Flow chart. Flow chart of enrolment of participants in the study

Fig. 2

Baseline plasma sST2 and Gal-3 levels. Comparison of baseline levels of sST2 (A) and Gal-3 (B) between controls and patients randomized in the study

Fig. 3

Baseline correlation between serum hs-TnI and plasma sST2 (A) and Gal-3 (B)

Fig. 4

Effects of intervention on plasma sST2, Gal-3 levels and serum hs-TnI. ∆%changes of sST2 (A), Gal-3 (B) and serum hs-TnI (C) in the liraglutide or lifestyle arm after achievement of the weight loss target. P-value: between-group difference in ∆%sST2, ∆%Gal-3 and ∆%hs-TnI values adjusted for baseline waist circumference, baseline triglycerides and baseline VAT

Fig. 5

Predictive role of Gal-3 levels. Correlation between baseline levels of Gal-3 with ∆beta-index in the lifestyle (A) and in the liraglutide arm (B). C ∆beta-index in the liraglutide and in the lifestyle arm in patients with baseline Gal-3 levels above or under the median

Fig. 6

Effect of intervention on plasma sT2 and Gal-3 levels during OGTT. Levels of sST2 and Gal-3 during OGTT in lifestyle and liraglutide arm before (A, C) and after (B, D) weight loss