Gliptins can exert favorable cardiovascular effects in experimental models. These changes, as an almost general rule, include improved endothelial function, reduction of inflammatory markers, oxidative stress, ischemia/reperfusion injury and atherogenesis. In addition, increased adiponectin levels and modest decreases in lipidemia and blood pressure were reported [
8]. Unfortunately, the favorable DPP4 inhibition features reported in experimental studies are not necessarily reflected in clinical studies [
7] and a definite relationship between gliptins treatment and improved CV outcomes remains uncertain and requires further clarification [
9,
14]. This could be due to the fact that in addition to the activation of incretin hormones, gliptins are multi-target compounds and may therefore affect various additional substrates, leading to undesirable effects [
15].
Effect of sitagliptin on CV function and biomarkers
Interestingly, the reduction in hs-CRP levels was independently associated with the reduction in HbA1c levels. A recent cross-over study showed that 6 weeks of treatment with sitagliptin significantly reduced the plasma levels of CRP and that there was a significant inverse correlation between changes in GLP-1 and CRP levels following sitagliptin therapy (r = 0.41, P = 0.01) [
16]. In the present study, sitagliptin did not significantly decrease hs-CRP levels (P = 0.0994), and the baseline hs-CRP level was not associated with ΔHbA1c (data not shown); therefore, the causal relationship between these factors remains unknown. However, these data suggest that there is a relationship between improvements in glucose–insulin homeostasis and low-grade inflammation that is somehow mediated by sitagliptin therapy in diabetic subjects with high CV risk.
Prolongation of the QT interval is an important risk factor for ventricular arrhythmia and sudden death. In the present study, the QT interval was significantly shortened from baseline to 12 months (P < 0.001), but there was no significant difference in the QT interval after correction for the change in HR (QTc interval) (P = 0.093). The basic research by Lee et al. showed that sitagliptin treatment could reverse calcium regulation in cardiomyocytes, which might contribute to its shortening effects on the QT interval in hypertensive animals [
17]. In other research, sitagliptin was not associated with prolongation of either the QT or QTc interval at clinically relevant concentrations in healthy individuals [
18]. Although the shortening of the QT interval in this study was mainly associated with the increased HR, the QTc interval still showed a trend toward shortening after sitagliptin therapy (P = 0.093). Thus, it is possible that sitagliptin therapy might lower the risks of ventricular arrhythmias and sudden death due to QT interval change, but further studies are clearly necessary to clarify whether benefits truly exist, particularly in patients with diabetes and comorbid CVD.
GLP-1 receptors are located in cardiac myocytes, coronary vascular endothelial cells, and smooth muscle cells; and, mainly in basic studies, GLP-1 has been reported to improve cardiac function and provide cardioprotective effects [
9]. Nikolaidis et al. reported that GLP-1 infusion in patients with acute myocardial infarction improved LV function compared with controls receiving standard care [
19]. However, in a randomized study of 80 diabetic patients with LV diastolic dysfunction, Oe et al. reported that sitagliptin for 6 months did not improve diastolic function [
20]. In the present study, in which we used patients with similar background characteristics, there was no significant change in BNP levels at 12 months, which is a reasonably sensitive maker for the cardiac dysfunction and development of HF. This was the same for the EF and septal e′ or E/e′, which reflect diastolic function or end-diastolic pressure, as the diagnostic and therapeutic index of HF. Endothelial dysfunction may also result from hyperglycemia and atherosclerosis, and some studies have suggested that sitagliptin can improve endothelial dysfunction in patients with type 2 diabetes and that gliptins may protect endothelial function through a GLP-1-dependent mechanism [
21,
22]. In contrast, Aoyagi et al. recently reported that gliptins attenuated endothelial function evaluated by flow-mediated vasodilatation [
23]. Also, sitagliptin has been shown to increase adiponectin levels [
24,
25], which could be associated with the anti-atherosclerotic effects. Although the measurement was not performed in the present study, sitagliptin treatment did not reduce markers of oxidative stress (urinary 8-OHdG) or inflammation (hs-CRP) but decreased HDL cholesterol levels. They did not affect endothelial function (RHI assessment) favorably. Thus, controversy remains around these aspects of treatment, and further investigation is needed.
Effect of sitagliptin on BP and albuminuria
We showed that sitagliptin reduced SBP and DBP independent of ΔHbA1c and that this was associated with a reduction in the level of albuminuria. It might also have been associated with a reduction in the LV mass index. Relatively few studies have examined the effects of DPP-4 inhibitors on BP in patients with diabetes [
26]. Presently, the effects are unclear and controversial in both clinical and basic studies, with some investigators reporting that gliptins may reduce BP [
27‐
29] and others reporting that these agents have no significant effects on BP [
30,
31]. Ogawa et al. reported sitagliptin treatment affected SBP but not DBP in 17 patients with hypertension; however, in that research, the degree of decrease in HbA1c levels did not significantly correlate with the degree of decrease in SBP (r = 0.24) [
28]. The results were, therefore, concordant with the present result in patients with mild diabetes but at high CV risk.
DPP-4 inhibition with sitagliptin increases the activity of GLP-1, which has been reported to decrease urinary salt uptake and increase urinary salt excretion [
32]. In fact, 6 months of treatment with exenatide, a GLP-1 receptor agonist, was associated with a greater reduction in SBP compared with placebo or insulin in pooled data from six trials including 2171 patients [
33]. The DURATION-2 study assessed the safety and efficacy of once weekly exenatide against maximum approved doses of sitagliptin, thiazolidinedione, or pioglitazone, and indicated that the reduction in SBP was significantly greater with exenatide once weekly than with sitagliptin in all patients after 26 weeks of treatment [
34]. The reason for the variation in the long-term effects of gliptins on BP remains unknown; however, it might be dependent on the patient’s physiological state because DPP-4 metabolizes both prohypertensive and antihypertensive peptides [
35]. Further clinical studies are necessary to elucidate the reasons for these differences. In the present study, HR significantly increased both in ECG analysis and ambulatory BP measurement. It might be associated with the reactive activation in the sympathetic nerve system or inhibition of the vagal nerve system to the BP decrease as suggested in the treatment of GLP-1 receptor agonists [
36]. However, the ΔHR was not associated with ΔSBP or ΔDBP at 12 months in the present study, and the other mechanism might be suggested. The point of the long-term effects or CV outcomes needs to be clarified with caution [
37].
The development of albuminuria is a key step in the progression of diabetic kidney disease, and worsening of albuminuria is a significant predictor of both progressive renal disease and CVD in patients with diabetes. The mechanisms underlying albuminuria are complex, and improved insulin sensitivity, weight loss, reduced blood glucose levels, lower BP, and suppressed inflammation might all contribute to a reduction in albuminuria. Several studies have reported a significant correlation between improvements of albuminuria and BP in patients with diabetes, hypertension, or nephropathy. Kawasaki et al. reported in a retrospective analysis that sitagliptin reduced albuminuria, and that this was independently associated with reductions in BP and eGFR in 247 patients with diabetes [
38]. In the present study, although sitagliptin did not reduce albuminuria overall, the reduction of BP was modestly associated with a decrease in albuminuria that was independent of ΔHbA1c. Groop et al. reported that treatment with linagliptin for diabetic patients with renal dysfunction led to a significant reduction in albuminuria, independent of changes in SBP [
39]. Although these findings may suggest direct renoprotective effects of DPP-4 inhibitors [
40], such a blood-pressure-independent effect was not observed in the present study. eGFR was deteriorated during the 12-month sitagliptin treatment, which was positively associated with ΔSBP or ΔDBP in the present study. Kawasaki et al. reported the similar findings and discussed that it might be related to the decrease in the glomerular internal pressure and be beneficial for the renal tissue [
38]. This mechanism resembles the decrease in eGFR observed through the action of renin–angiotensin system inhibitors in decreasing glomerular internal pressure and albuminuria. Further evidence is needed to confirm these renal effects, particularly in the long-term and explore the underlying mechanisms in patients with diabetes [
41].
Study limitations
First, this was an uncontrolled study and the effects of sitagliptin were only assessed by comparison of measurements at three independent time points. It might be possible that the significant changes in several measures were affected by factors other than sitagliptin treatment. However, all the patients were educated about the importance of diet and physical exercise, and most had stable medical treatment at least 1 month before the study inclusion. Second, the study assessments took place at 3 or 12 months, and may have been too short for the assessment of CV effects and outcomes. Therefore, the analysis may have been insufficient to provide conclusive evidence for long-term CV effects and outcomes associated with sitagliptin. Future studies must examine the BP lowering effects of sitagliptin and its direct influence on CV outcomes. Third, most of the participants had early-stage diabetes (69.3 % had an HbA1c <7.0 %) and higher age (73.4 % were aged >65 years), meaning that our results cannot be extrapolated to all patients with diabetes and high CV risk. Finally, our study design also led to the exclusion of some patients with severe chronic kidney disease or end-stage renal disease, which limits the applicability of the data in these settings.