Dipeptidyl peptidase inhibitors DPP is
Dipeptidyl peptidase-4 inhibitors (DPP-4is or gliptins) stimulate both post-meal insulin secretion and inhibit Phusion high-fidelity DNA polymerase secretion in a glucose-dependent manner, rendering the risk of hypoglycaemia minimal . They are weight-neutral or may even induce slight weight loss. Their good efficacy has been demonstrated in clinical placebo-controlled studies as well as in head-to-head randomized controlled trials (RCTs) comparing DPP-4is with other active glucose-lowering agents , , especially SUs , , while DPP-4i CV safety has been proven in three large prospective RCTs showing non-inferiority vs placebo in T2DM patients with CV disease , , . However, there are as yet no prospective RCTs available that compare CV outcomes with DPP-4is or other active agents, especially SUs . Nevertheless, DPP-4is are currently recommended as second- or third-line therapies for the management of T2DM as an alternative to SUs , , , and have even been preferred to SUs in some recent US  and French  guidelines.
To identify relevant studies, an extensive literature search of MEDLINE and Embase was performed from January 2005 to February 2018, using the terms ‘DPP-4 inhibitor’ or ‘gliptin’ or ‘sulphonylurea’, or ‘sulfonylurea’ combined with ‘cardiovascular’ or ‘myocardial infarction’ or ‘stroke’ or ‘mortality’. No language restrictions were imposed. Reference lists of original studies, narrative reviews and previous systematic reviews were also carefully examined. Despite the lack of an available head-to-head CV outcome trial (CVOT), useful information may yet be derived from meta-analyses comparing CV outcomes separately for each pharmacological class. However, in a recent extensive meta-analysis comparing clinical outcomes and adverse events associated with glucose-lowering drugs in patients with T2DM, no data were reported for CV outcomes in patients treated with SUs compared with DPP-4is .
CV outcomes in clinical studies: separate analyses within each pharmacological class
CV outcomes in studies of head-to-head comparisons
Are all SUs and DPP-4is similar for CV safety? When different agents belong to a same pharmacological category, the question that arises is whether a class effect may be involved in either efficacy or safety issues. Indeed, the main criteria for considering a class effect or not have recently been discussed, taking CV outcomes with GLP-1 receptor agonists as an example .
Disclosure of interest
Introduction Diabetes is associated with alterations in the retinal microvasculature, including morphological endothelial changes. The retinal microvessels consist of a continuous endothelium and intramural pericytes in the basal lamina [, , ]. Chronic hyperglycemia leads to histopathological changes that can be detected in the early stages of diabetic retinopathy, such as capillary basement membrane thickening, intraretinal hemorrhages, development of microaneurysms, hard exudates, dilation of retinal veins and tortuosity of capillaries. Nonproliferative diabetic retinopathy stage is accompanied by focal capillary and arterial occlusion that can result in retinal ischemia, promoting therefore the production of angiogenic factors that contribute to the formation of new fragile vessels, the hallmark of the proliferative diabetic retinopathy. If untreated, hemorrhages into the vitreous or retinal detachment with consequent vision loss can also occur [, , , ]. It is well established that diabetic retinopathy has many features of a low-grade chronic inflammatory condition, such as increased vascular permeability, infiltration of immune cells, increased expression of cytokines and chemokines, and microglia activation . Enhanced expression and secretion of pro-inflammatory cytokines, including tumor necrosis factor (TNF)-α, have been detected in the vitreous of diabetic patients [9,10]. In addition, TNF-α is known to be implicated in diabetic retinopathy progression since it promotes the adhesion of leukocytes to the retinal endothelium and blood-retinal barrier (BRB) permeability [, , , , , , ].