catherine only

Contents lists available at ScienceDirect

Pharmacological Research

journal homepage: www.elsevier.com/locate/yphrs

Antidiabetic drugs and blood pressure changes

Ioannis Iliasa,1, Costas Thomopoulosb,*,1, Helena Michalopoulouc, George Bazoukisd, Costas Tsioufise, Thomas Makrisb

a Department of Endocrinology, Helena Venizelou Hospital, Athens, Greece bDepartment of Cardiology, Helena Venizelou Hospital, Athens, Greece c Department of Cardiology, Metaxa Cancer Hospital, Piraeus, Greece d Second Department of Cardiology, Evangelismos Hospital, Athens, Greece e First Cardiology Clinic, Hippokration Hospital, Athens University, Athens, Greece

A R T I C L E I N F O

Keywords: Antidiabetic drugs Type 2 diabetes mellitus Randomized clinical trials Blood pressure reduction Meta-analysis

A B S T R A C T

New era antidiabetic drugs are characterized by cardiovascular safety, including specific outcome benefits ob- served in randomized clinical trials (RCTs). It has been postulated that the favorable effects of new antidiabetic agents are related both to better control of blood pressure (BP) levels and to activation of multiple anti-ather- osclerotic properties. In this review, we aimed to assess whether antidiabetic drugs have a pressor effect in glucose control and outcome-oriented RCTs, and to summarize the activated pathophysiological mechanisms relevant to BP control following the use of different antidiabetic drug classes. We also tried to determine which, if any, are the BP-lowering effects of more intense vs less intense glucose-lowering strategy irrespectively of trial antidiabetic regimen. To provide more robust results and evidence-based argumentation, a meta-analysis of placebo-controlled antidiabetic drug RCTs was undertaken to estimate the ongoing BP reduction for all con- sidered and each separate drug class alone. This quantitative synthesis might be helpful for the clinician 1) to select or avoid the use of some classes of antidiabetic agents with a potential favorable or adverse pressor effect, respectively 2) to organize the overall drug regimen in patients with diabetes mellitus and minimize side effects because of concomitant use of drugs with established pressor effect (i.e. antihypertensive agents). This review was also organized to indicate whether BP change associated with different antidiabetic treatments may explain the specific macrovascular outcome benefits. Between all antidiabetic drugs including exogenous insulin, only sodium-glucose cotransporter 2 inhibitors produce a clinically important BP-lowering effect, but this BP re- duction alone cannot explain the observed cardiovascular benefit.

1. Introduction

Type 2 diabetes mellitus (T2DM) and hypertension are comorbid clinical conditions that interact to create a variable-degree vascular deterioration, and thus increasing the risk of macrovascular disease [1]. The combined management of diabetes mellitus and hypertension, through blood glucose and blood pressure (BP) reduction is of clinical priority in the treatment of patients with diabetes because it can reduce

the burden of incident major cardiovascular events and microvascular complications (including the development of chronic kidney disease) [2]. Although, earlier antidiabetic agents (insulin, sulfonylureas, met- formin, and thiazolidinediones [TZDs]) were found to constantly re- duce microvascular complications [3], their effect on major cardio- vascular events was not beneficial, possibly because the studies were underpowered to demonstrate changes on macrovascular complications within a conventional trial period of less than five years [4]. Concerns

https://doi.org/10.1016/j.phrs.2020.105108 Received 16 May 2020; Received in revised form 30 May 2020; Accepted 24 July 2020

Abbreviations: ACCORD, action to control cardiovascular risk in diabetes; ADVANCE, action in diabetes and vascular disease PreterAx and diamicron MR controlled evaluation; Alecardio, aleglitazar on cardiovascular outcomes; BP, blood pressure; CANVAS, canagliflozin cardiovascular assessment study; CAROLINA, cardio- vascular outcome study of linagliptin vs glimepiride in type 2 diabetes; DPP4, dipeptidyl-peptidase 4; FDA, food and drug administration; GLP1, glucagon-like peptide-1; HOME, the hyperinsulinemia the outcome of its metabolic effects; ORIGIN, outcome reduction with an initial glargine intervention; PPAR, peroxisome proliferator-activated receptor; PROACTIVE, prospective pioglitazone clinical trial in macrovascular events; RCTs, randomized clinical trials; SGLT-2, sodium-glucose cotransporter 2; SGLTs, sodium-glucose co-transporters; SPRINT, systolic blood pressure intervention trial; TZDs, thiazolidinediones; T2DM, type 2 diabetes mellitus; UKPDS-33, UK prospective diabetes study 33

⁎ Corresponding author. E-mail address: thokos@otenet.gr (C. Thomopoulos).

1 The first two authors equally contributed to this work.

Pharmacological Research 161 (2020) 105108

Available online 30 July 2020 1043-6618/ © 2020 Elsevier Ltd. All rights reserved.

T

about the cardiovascular safety of rosiglitazone led initially the United States Food and Drug Administration (FDA) in 2008 to mandate that new antidiabetic agents be tested for cardiovascular safety, thus re- quiring much larger outcome trials [5,6]. Newer antidiabetic drugs (dipeptidyl-peptidase 4 [DPP4] inhibitors, glucagon-like peptide-1 [GLP1] receptor agonists, and sodium-glucose cotransporter-2 [SGLT2] inhibitors) were tested in double-blind placebo-controlled randomized clinical trials (RCTs) with neutral and occasionally beneficial effects compared to their placebo counterparts [7,8].

Because outcome RCTs of the new antidiabetic agents were not designed to produce differential glucose lowering between randomized groups (Table 1) [9–34], as measured by glycated hemoglobin levels, it is by and large undetermined which factors may have contributed to their cardiovascular safety. It has been hypothesized that differences in BP levels during follow-up may have, at least in part, determined a differential outcome incidence between active and placebo arm in large antidiabetic treatment RCTs [35,36]. The purpose of the present

review, was to examine whether different antidiabetic treatment classes may have a substantial pressor effect against placebo, and to evaluate the potential impact of BP change as a modulator of cardiovascular risk reduction in antidiabetic treatment outcome RCTs. To quantify the BP- lowering effect of different antidiabetic drugs we updated our previous comprehensive meta-analysis of RCTs in T2DM [35].

2. Trial selection and effect of glucose-lowering on various outcomes

We selected RCTs of different antidiabetic drugs vs placebo or more intense vs less intense blood glucose-lowering treatment (Table 1) re- porting enough data for baseline and/or attained systolic/diastolic BP difference between groups. Trial eligibility has been previously ex- plained in detail [35]. In brief, we included RCTs of antidiabetic drugs in which the attained glycated hemoglobin difference was at least 0.2 % between arms; at least 5 outcomes during follow-up; duration of at least 12 months. However, in this present work we included two more RCTs that have been recently published [22,27]. Trials of patients with heart failure in which outcome incidence may be hindered by BP-lowering were not considered. A random-effects model was by default selected for all analysis, while we used the attained systolic and diastolic BP together with the respective standard deviations from each selected trial, to assess the mean attained BP difference and the corresponding 95 % confidence intervals. The proportion of inconsistency across stu- dies not explained by chance was assessed by the I-squared index. Summary outcome risk ratios were calculated as previously described [35].

In this updated analysis (26 trials; 186,565 T2DM patients, follow- up, 3.55 years), the outcome risk reduction of five prespecified fatal and non-fatal outcomes is presented in Supplementary Table 1. It can be appreciated that blood glucose reduction compared to placebo or less intense glucose-lowering (attained mean glycated hemoglobin differ- ence of -0.5 %) was associated with a significant reduction of coronary heart disease, stroke, cardiovascular death and all-cause mortality. This beneficial effect was also associated with ongoing systolic BP reduction of -1.4 mmHg, while diastolic BP was not reduced (Fig. 1).

3. Insulin

Patients with T2DM are characterized by insulin resistance (i.e. defective insulin action at cellular level) and beta-cell dysfunction, while hypertensive patients are expected to have sympathetic over- activity and various degrees of vascular damage, ranging from en- dothelial dysfunction to overt atherosclerotic disease. The association between T2DM and hypertension has been observed in different clinical studies but this association is confounded by obesity [37]. Obesity is associated with increased blood volume and cardiac output in condi- tions of decreased vascular resistance because of the adipose tissue vascular bed enlargement. Consequently, increased body adiposity is not constantly associated with BP elevation. Although, the association between diabetes mellitus and hypertension remains present after ad- justment for obesity, it is suggested that the common pathophysiolo- gical modulating background of any observed BP increase stems from insulin resistance at skeletal muscle level [38]. Insulin at physiological concentrations increases limb blood flow by stimulating nitric oxide synthase but also by enhancing acetylcholine-mediated vasodilation. In patients with insulin resistance (patients with T2DM and/or hyperten- sion) vasodilation - in response to increased plasma levels of insulin - is reduced, independently of previously impaired or not endothelial function [38].

Insulin resistance is associated with hyperinsulinemia, given that cellular ability to enter glucose within the cell, as a response to the available insulin, is not preserved. Hyperinsulinemia not only impairs the glucose pathway, but may stimulate other intracellular pathways, such as the growth signaling cascade that may lead to cellular

Table 1 Blood pressure-lowering effect of different antidiabetic drugs in outcome ran- domized clinical trials.

Outcome RCTs by drug [REF]

Achieved mean SBP/DBP Difference, mmHg

Achieved mean body weight difference, Kg

Achieved HBA1c reduction, %

Insulin ORIGIN [9] 0/0 2.1 −0.4 Third generation SUs ADVANCE [10] −2.4/-0.8 1.1 −0.7 Metformin HOME [11] −0.5/-0.9 −5 −0.4 TZDs PROACTIVE [12] −3/0 3.2 −0.5 ALECARDIO [13] −2.2/-0.7 3.3 −0.6 DPP-4 inhibitors CARMELINA [14] −0.2/0 0 −0.4 EXAMINE [15] −1.3/-0.1 0.3 −0.3 SAVOR-TIMI 53

[16] 0/0 −0.5 −0.3

TECOS [17] −0.5/-0.1 −0.2 −0.3 GLP-1 receptor

Agonists ELIXA [18] −0.8/0 −0.7 −0.3 EXSCEL [19] −1.6/0.3 −1.3 −0.5 LEADER [20] −1.2/-0.6 −2.1 −0.4 HARMONY-

outcomes [21] −0.7/-0.1 −0.7 −0.6

REWIND [22] −1.7/0.1 −0.5 −0.6 SUSTAIN-6 [23] −2.1/-0.1 −3.6 −0.9 SGLT-2 inhibitors CANVAS-program

[24] −3.9/-1.4 −1.6 −0.6

DECLARE-TIMI 58 [25]

−2.7/-0.7 −1.8 −0.4

EMPAREG-outcome [26]

−4/-1 −1.8 −0.5

CREDENCE [27] −3.3/-1 −0.8 −0.3 M vs L* ACCORD [28] −1/-1 3.1 −1.1 ADVANCE [10] −2.4/-0.8 1.1 −0.7 BARI-2D [29] −0.2/-0.3 −0.8 −0.4 Kumamoto [30] −1/-2 0.2 −2.4 SURE [31] −2/-3 NR 1 −0.7 UKPDS-33 [32] 0.8/0.4 NR −0.9 VA-CSDM [33] −0.1/0 4.1 −2.1 VADT [34] 2/-1 −1.5

DBP, achieved diastolic blood pressure; DPP, dipeptidyl-peptidase; glucagon- like peptide, GLP; SGLT, sodium-glucose co-transporters; HBA1c, glycated he- moglobin; NR, not reported; SBP, achieved systolic blood pressure; SUs, sulfo- nylureas; TZDs, thiazolidinediones. Minus sign indicates that drug reduces more BP or body weight compared to placebo or less intense glucose-lowering treatment. *, more vs less intense glucose-lowering strategy. Note: for trial acronyms please refer to reference list.

I. Ilias, et al. Pharmacological Research 161 (2020) 105108

2

proliferation and consequent loss of vessel autoregulation [39]. Hy- perinsulinemia represents the central effector modulating the home- ostasis of glucose metabolism and BP levels, not only in hypertensive but also in normotensive states. Longitudinal evidence suggest that hyperinsulinemia may precede the development of hypertension, while the reverse temporal sequence has not been documented [38].

The multifactorial pathophysiological cross-talk between pharma- cological hyperinsulinemia and BP levels might be bi-directional be- cause, for example, high insulin levels activate the adrenergic nervous system and conversely adrenergic overactivity may result in hyper- insulinemia via insulin resistance [39]. The volume-mediated BP ele- vation following insulin treatment seems independent of hypertension and hemodynamic parameters, such as glomerular filtration rate [38]. However, the concomitant activation of the sympathetic system may accentuate the sodium-retention properties of insulin. Furthermore, insulin-induced hypokalemia through activation of sodium-potassium cellular channels increases plasma renin and angiotensin II levels while decreasing the serum aldosterone concentration. The interplay between the physiological actions of insulin (stimulation of nitric oxide synthase, insulin-induced vasodilation) and the adverse consequences of insulin resistance on the vasculature and at cellular level (e.g. increased so- dium retention, sympathetic overactivity, hypokalemia, mitogen acti- vated protein kinase) may be reciprocally neutralized or result in in- creased vascular stiffness, volume overload, electrolytic disturbances and BP increase. In these lines, the effect of exogenous insulin in pa- tients with diabetes mellitus and/or hypertension may have a variable BP response, also depending on the natural history of the disease.

Insulin treatment initiation in middle aged overweight patients with T2DM was associated with a significant BP elevation compared with controls, and a difference in systolic/diastolic BP of 7.2/5.8 mmHg, respectively, was observed between the two groups [40]. Of note, weight was significantly increased during follow-up in the group of patients receiving insulin, and the difference in weight between groups at the last visit was 5.2 Kg. Since obesity is associated with a higher degree of insulin resistance, it can be hypothesized that the introduc- tion of exogenous insulin probably deteriorates the pressor effects of insulin resistance. In another study [41], with a cross-over design, in- cluding obese patients with essential hypertension but without diabetes mellitus, the administration of insulin for two weeks was associated with post-intervention office systolic/diastolic BP reduction (-3.8/- 3.3 mmHg), which was also confirmed by ambulatory systolic/diastolic BP reduction (-1.1/-1.2 mmHg). This effect was found to be mediated by better insulin-induced vasodilation (maximal leg blood flow during plethysmography) and insulin sensitivity (hyperinsulinemic, eu- glycemic glucose clamp). However, the cross-sectional design and

limited number of participants could not generate definitive answers about the BP-lowering effects of insulin suggested by this clinical model.

In the Outcome Reduction with an Initial Glargine Intervention (ORIGIN) trial [9], insulin glargine or standard care were randomly administered in patients with impaired fasting glucose or tolerance or T2DM (hypertension prevalence, 80 %; 10-year fatal cardiovascular risk, 14.8 %). For a weight difference of 2.1 Kg, the ongoing systolic/ diastolic BP between groups remained identical during a 6-year period. However, the frequency and modality of BP measurements in the ORIGIN trial remains unknown, which makes any conclusion about BP changes at least biased. To conclude whether exogenous insulin may have a pressor effect in patients with T2DM or other insulin resistance states, including essential hypertension is problematic. Insulin sensi- tivity and not hyperinsulinemia may represent the “missing link” be- tween insulin resistance states and BP elevation [38]. At present, based on the ORIGIN trial [9], the clinical effect of insulin on BP levels might be considered neutral. Of course, no meta-analysis for insulin-based treatment was performed because no other trial with ongoing BP-dif- ferences between insulin and placebo has been conducted so far.

4. Sulfonylureas

This class of drugs stimulates insulin secretion from pancreatic beta- cells by inhibiting potassium efflux and, in a following step, decreases hepatic insulin clearance. However, sulfonylureas are a very hetero- geneous category with the first-generation drugs (e.g. chlorpromamide) not being currently used due to increased rate of side effects. Second and third-generation agents (e.g. glibenclamide, glipizide, gliclazide, glimepiride etc.) are more effective at lower therapeutic doses with less side-effects compared to the first-generation agents. Antidiabetic treatment with sulfonylureas is associated with hyperinsulinemia, ac- tivation of the sympathetic nervous system and inhibition of the po- tassium-dependent adenosine triphosphate channel, which separately or in combination may increase vascular tone, reduce the vasodilatory activity and increase BP levels. Although, the extra-pancreatic detri- mental effects of sulfonylureas are mediated by activation of myo- cardial or vascular receptors for example, third-generation agents, like gliclazide, act only on the pancreatic receptor and potential effects on BP levels may be mediated by improvement of tissue insulin sensitivity [42].

In the UK Prospective Diabetes Study (UKPDS)-33 [32], after a 6- year period, treatment with a first-generation sulfonylurea, chlorpro- pamide, was associated with increased ongoing systolic/diastolic BP by 5/2mmHg compared to other treatment allocations including

Fig. 1. Achieved systolic and diastolic blood pressure difference in antidiabetic drug outcome RCTs. TZDs, thiazolidinediones; DBP, diastolic blood pressure; DPP-4; dipeptidyl peptidase; GLP1, glucagon-like peptide-1; SBP, systolic blood pressure; SGLT2, sodium-glucose cotransporter 2; M vs L, more vs less. The minus sign indicates a lower BP value in the first group, while black color indicates systolic BP reduction and grey color diastolic BP reduction.

I. Ilias, et al. Pharmacological Research 161 (2020) 105108

3

conventional treatment (glibenclamide, a second-generation sulfony- lurea, and insulin). Among sulfonylureas, at present, gliclazide is the most prescribed one, since a large amount of observational data showed its cardiovascular safety [42]. Moreover, in the ADVANCE (Action in Diabetes and Vascular disease: PreterAx and Diamicron MR Controlled Evaluation) trial [10] there was a significant reduction in microvascular complications in the gliclazide-treated intensive arm compared to the non-gliclazide-treated-conventional arm. This effect was associated with systolic/diastolic attained BP reduction of -2.4/-0.8 mmHg. This may suggest that third generation agents have a more favorable pressor effect compared to earlier generation sulfonylureas. Finally, in the Cardiovascular Outcome Study of Linagliptin vs Glimepiride in Type 2 Diabetes (CAROLINA) trial [43], glimepiride, a third-generation sulfo- nylurea was tested against linagliptin, a DPP-4 inhibitor, in T2DM pa- tients without background insulin treatment. Both treatments were associated with no differential incidence on the composite cardiovas- cular outcome, while no BP difference throughout the follow-up period was detected between the two drugs. The currently used third-genera- tion sulfonylureas have a no significant pressor effect. No outcome RCTs-based meta-analysis was conducted because of scarce available data.

5. Metformin

The BP-lowering effect of metformin has been by and large ques- tioned [44]. Different pathophysiological mechanisms, such as body weight and insulin resistance reduction, attenuation of insulin-medi- ated vasoconstriction, adrenergic receptor deactivation, reduction of intracytoplasmic calcium, inhibition of sympathetic overdrive (espe- cially in high-sodium intake dietary patterns), increase of glomerular filtration rate and sodium excretion and improvement of endothelial function, have been proposed in experimental in-vivo and ex-vivo stu- dies as potential contributors to the BP-lowering effect of metformin. However, because of different experimental designs used across studies (e.g. mode of BP measurements, extent of BP elevation, route of drug administration) metformin has not been constantly associated with a BP reduction [44].

The clinical research about the BP-lowering effects of metformin was summarized in a meta-analysis of 26 placebo-controlled RCTs in patients without diabetes mellitus, showing a reduction of systolic BP by almost 2mmHg, whereas diastolic BP was not reduced in the same setting [45]. This trivial systolic BP-lowering effect of metformin was not unexpected since hypertensive patients across the selected studies were underrepresented. Indeed, previous experimental evidence sug- gested that metformin reduces BP levels in hypertensive and not nor- motensive animal models [44]. Another possible explanation for this small BP-lowering effect is that metformin was tested in heterogeneous clinical conditions, including individuals with polycystic ovary syn- drome, schizophrenia, impaired glucose tolerance, hypertension etc., in which the main investigational question was different from the as- sessment of BP response and the modality of BP measurement in many studies remains largely obscure. In another meta-analysis including 21 T2DM RCTs [46], systolic and diastolic BP were not reduced following metformin administration. Overall, the evidence provided by met- formin RCTs is highly biased because of selection bias across studies, the variable mode of BP measurements, the unequal background anti- hypertensive treatment and the high rate of individuals with normal BP levels being recruited in different trials. Future clinical trials should focus on the differential effects of metformin on BP by discriminating hypertensive from normotensive patients, as well patients with diabetes mellitus from those at risk for diabetes mellitus. Measurement of dietary salt at baseline and follow-up (e.g. 24-h urinary sodium excre- tion) might also be important in combination with office and/or am- bulatory BP measurements, at different times during follow-up [44]. Finally, the Hyperinsulinemia: The Outcome of its Metabolic Effects (HOME) trial [11] reported that the systolic/diastolic BP difference

between metformin and placebo was less than 1mmHg (i.e.−0.5/- 0.9 mmHg, respectively) in low-moderate cardiovascular risk patients with T2DM having a hypertension prevalence of almost 40 %. Based on the above limited evidence, the BP-lowering effect of metformin is very limited and clinically not significant.

6. PPAR (peroxisome proliferator-activated receptor)-gamma agonists/TZDs

Their action is mediated through a nuclear receptor, namely per- oxisome proliferator-activated receptor, which regulates the expression of various genes involved in adipocyte cell differentiation and pathways of lipid and glucose metabolism. This particular category of drugs may improve tissue sensitivity to insulin, and reduce insulin resistance. However, the popularity in prescribing this category of antidiabetic agents has declined due to increased rate of heart failure in high risk patients [5]. Beyond blood glucose reduction, PPAR-gamma ligands may modulate - by transcriptional control at both the mediator and receptor levels – the activity of the renin-angiotensin system. PPAR- gamma receptors are expressed in two distinct renal areas: at the jux- taglomerular apparatus level, in which they promote the synthesis of renin and at the distal collecting ducts, and therefore may promote aldosterone urinary excretion, and water/sodium reabsorption [47]. The hypertensive effect of an activated renin-angiotensin system is at- tenuated by PPAR-gamma ligands since they reduce angiotensin II, angiotensin converting enzyme 1, aldosterone production and angio- tensin receptor 1 gene expression. Although PPAR-gamma stimulation offsets aldosterone action (decreased production at adrenal cortex and increased urinary elimination), it intensifies antidiuresis at the level of distal collecting tubule. The initial blood volume overload may supply with fluids the interstitial compartment leading to peripheral edema development. Why this latter phenomenon happens is not fully eluci- dated, however, individuals with single nucleotide polymorphisms in the beta-1 adrenergic receptor gene may develop peripheral edema more frequently during PPAR-gamma agonism compared with patients without this specific mutation [48]. In patients with peripheral edemas, stroke volume is reduced, and thus BP values are not elevated. In a meta-analysis of RCTs of glucose control with TZDs, systolic/diastolic BP was reduced by -4/-2 mmHg as a change from baseline BP levels. In the Prospective Pioglitazone Clinical Trial in Macrovascular Events (PROACTIVE) [12], which included high-risk T2DM patients with a baseline hypertension prevalence of 76 %, pioglitazone reduced only systolic BP by 3mmHg compared to controls whereas no reduction in diastolic BP was noted. In the Aleglitazar on Cardiovascular Outcomes (AleCardio) study [13], the dual PPAR agonism of aleglitazar having a balanced affinity for the PPAR-alpha and PPAR-gamma subtypes was associated with a small systolic/diastolic BP reduction compared to placebo by -2.2/-0.7 mmHg in T2DM patients after an acute coronary syndrome. Again, based on limited evidence from large outcome RCTs, TZDs reduce in a small extent only systolic BP (Fig. 1). Of note that this small BP-lowering was not accompanied by various outcome risk re- duction, whereas heart failure incidence was increased by 32 % com- pared to placebo (Supplementary Table S1).

7. DDP4 Inhibitors

DDP4 inhibitors are incretins and lower blood glucose by increasing insulin release at the pancreatic level. The use of DDP4 inhibitors is not accompanied by weight gain compared with older antidiabetic drugs and are increasingly used as an adjunct to first-line therapy with met- formin. However, this particular category may precipitate heart failure in patients at increased cardiovascular risk or may aggravate the clin- ical course of patients with pre-existing left ventricular dysfunction [49]. The detrimental effects of DDP4 inhibitors, however, might not become clinically important if DDP4 inhibitors were to exert mean- ingful natriuretic effects, and thus reducing the cardiac preload. There

I. Ilias, et al. Pharmacological Research 161 (2020) 105108

4

is experimental evidence suggesting that when insulin signaling is po- tentiated by DDP4 inhibitors, adverse cardiovascular adaptations may occur and sodium handling at the kidney level may be potentiated [49]. By contrast, another line of evidence indicates that DDP4 inhibitors can also have a direct natriuretic effect, acting on the distal renal tubules, but this effect on sodium handling is modest. This latter phenomenon is mediated by potentiation of the stromal cell–derived factor-1 and not through a GLP-1 receptor interplay. Thus, at the kidney level, DDP4 inhibitors may have two opposite actions: sodium retention mediated by hyperinsulinemia versus mild natriuresis at nephrons’ distal tubule. Furthermore, DDP4 inhibitors may promote systemic vasodilation by cleavage of endogenous natriuretic factors, like neuropeptide Y1 and substance P, as well as through increased nitric oxide bioavailability. But how is this experimental information is translated to changes in BP in patients with T2DM receiving DDP4 inhibitors? A meta-analysis considered six RCTs in T2DM patients (n=1500) in which DDP4 in- hibitors were tested against placebo or no treatment [50]. The mean change of systolic/diastolic BP from baseline was -3.0/-0.5 mmHg. However, this type of analysis has the problem that changes from baseline depend on the initial BP levels and that small studies are characterized by unequal between-arms BP levels. In our present meta- analysis including four large outcome RCTs (n= 43,522 patients) [14–17] we considered the comparison between DDP4 inhibitors and placebo. The ongoing systolic/diastolic BP reduction with DDP-in- hibitors was trivial over a 2.3-year period (Fig. 1). The current evidence does not grant any clinically important BP-lowering effect of DDP-4 inhibitors according to the results of RCTs focused on anti- hyperglycemic effect or cardiovascular safety issues of this category. This should be seen in the context of the neutral effect of this drug class on fatal and non-fatal cardiovascular outcomes (Supplementary Table S1).

8. GLP-1 agonists

Endogenous GLP-1 is an incretin hormone that stimulates insulin production and inhibits glucagon secretion in a dose-depending fashion modulated by blood glucose levels [8]. Other accessory physiological effects of GLP-1 are the inhibition of gastric emptying, and the reduc- tion of appetite and food intake. Receptors for GLP-1 are distributed beyond the pancreas to the autonomic nervous system, blood vessels, sinoatrial node and Brunner’s intestinal glands. Several hypotheses have been raised to explain a potential BP-lowering effect of GLP-1 agonists. However, the experimental evidence supporting a BP-low- ering effect was not uniform, especially when administered acutely in animal models. Their chronic administration in both animals and hu- mans was accompanied by a trivial reduction in systolic but not in diastolic BP [51]. Several mechanisms have been proposed. At the vascular level, experimental evidence in humans raised the hypothesis that GLP-1 activation may have vasodilating properties by mechanisms both dependent and independent of nitric oxide stimulation, especially in those with impaired glucose metabolism, but not in healthy in- dividuals. This phenomenon is possibly associated with insulin-medi- ated enhancement of vascular reactivity, which is linked to improve- ment of tissue insulin sensitivity. Additional mechanisms by which GLP- 1 receptor agonists may be associated with ameliorated vascular per- formance and chronic BP-lowering effects are the reduction of in- tracellular calcium, the antiproliferative actions on smooth muscle cells and the reduction of systemic inflammation. The diuretic and na- triuretic action of GLP-1 receptor agonists might be mediated by in- hibition of the sodium-hydrogen exchanger-3 (located at the brush border of the renal proximal tubule). However, the acute administra- tion of GLP-1 receptor agonists likely increases the effective renal plasma flow and glomerular filtration rate, at least in healthy in- dividuals, resulting from an increase in heart rate and cardiac output [51]. In a meta-analysis of 14 placebo-controlled glucose control RCTs, only systolic BP was modestly reduced by -1.6mmHg, while diastolic

BP remained unaltered [52]. In the present synopsis of six large pla- cebo-controlled outcome RCTs [18–23], including T2DM patients (n= 52,821), treatment with a GLP-1 agonist compared to placebo reduced the achieved systolic BP by -1.3mmHg over a mean period of 3.2 years (Fig. 1). Although the neutral effect on diastolic BP is con- sistently seen in glucose control and outcome-oriented RCTs, the dif- ferent extent of systolic BP-lowering within the GLP-1 receptor agonists’ class may be related to intrinsic properties of different agents, clinical characteristics of each study population, mode of administration and multiple background medications. Weight loss may also have a con- founding potential on the BP-lowering effect of GLP-1 receptor agonists, but notably, BP reduction precedes any body weight reduction. The overall evidence suggests that GLP-1 receptor agonists may produce a trivial systolic BP reduction of less than 2mmHg, but the prevailing mechanism supporting this subtle reduction yet remains unclear. Blood glucose and systolic BP-lowering of -0.5 % (in terms of attained gly- cated hemoglobin) and -1.3mmHg, respectively, was accompanied by significant reduction of fatal outcomes and stroke (Supplementary Table S1).

9. SGLT-2 inhibitors

Sodium-glucose co-transporters (SGLTs) are partially responsible for blood glucose entry to cells across a sodium gradient, which is actively maintained by the sodium-potassium adenosine triphosphate pump. At the renal level, SGLTs are mediating reabsorption of the filtrated glu- cose at the glomeruli level back to the bloodstream [7]. Among the two known SGLTs, the SGLT-2 subtype is expressed at the initial convoluted portion of the proximal tubule and can manage 90 % of the filtrated glucose; this effect is at variance with the SGLT-1 subtype, expressed more distally, which helps the kidney to reabsorb the remaining 10 % of filtrated glucose. Interestingly, the expression of SGLT-2 at kidney level is significantly higher in patients with T2DM compared to healthy in- dividuals. SGLT-2 inhibition is associated with reduced glucose and sodium reabsorption. This increased rate of osmotic diuresis and na- triuresis may be the limiting step for BP reduction observed in SGLT-2 inhibitor RCTs. However, osmotic diuresis and increased natriuresis are associated with reduced renin production from the juxtaglomerular apparatus, vasoconstriction of the afferent and vasodilation of the ef- ferent arteriole at the glomeruli level, subsequent reduction of the fil- tration rate and intraglomerular hydrostatic pressure. A previous meta- analysis [53], including a large number short-term duration (< 1 year) placebo-controlled RCTs, demonstrated that systolic/diastolic BP change from baseline was higher in SGLT-2 inhibitor-treated patients as compared with placebo-treated group by -3.8/-1.5 mmHg. However, this BP-lowering effect was accompanied by significant change in body weight and hematocrit (-1.7 Kg and +2.4 %) as compared with baseline levels. Whether weight change and hemoconcentration contribute to BP-reduction or simply represent epiphenomena of the osmotic diuresis remain unresolved. In our present meta-analysis, based on four large placebo-controlled RCTs [24–27] in T2DM patients (n= 38,723), the achieved systolic/diastolic BP difference of -3.5/-1 mmHg (Fig. 1) over a mean follow-up period of 3.7 years. SGLT2-inhibitors lowered gly- cated hemoglobin by -0.46 % compared to placebo and this blood glucose lowering was associated with risk reduction of all fatal and non- fatal outcomes expect stroke (Supplementary Table S1).

The finding that BP reduction at short-term is continued over a period of almost four years, together with the finding that increased urinary volume returns to pre-treatment levels after a 3-month period may suggest that osmotic diuresis is not the only mechanism inducing BP reduction. Other proposed mechanisms include nephron re- modeling, reduction of arterial stiffness, vascular smooth muscle-cell relaxation secondary to negative sodium balance, reduction of sympa- thetic drive, and glycosuria-mediated weight loss independent of fluid contraction [54].

Overall, SGLT-2 inhibitors may produce a moderate but clinically

I. Ilias, et al. Pharmacological Research 161 (2020) 105108

5

important systolic and diastolic BP-lowering of almost 4 and 1mmHg, respectively, principally driven by diuresis. This BP-lowering effect appear not be accompanied by increased rate of hypotensive phe- nomena, like syncope in T2DM patients with hypertension already under optimal antihypertensive treatment. However, special attention should be paid in hypertensive patients under traditional diuretic agents because of increased rate of volume depletion [24] or acute kidney injury [25,26] noted with SGLT2s.

10. More vs less intense antidiabetic treatment

In our present meta-analysis, the BP-lowering effect of combined antidiabetic treatment strategies aiming to intentionally produce a differential blood glucose-lowering between randomized arms was tested in 8 large RCTs (n=28,314; follow-up, 5.1 years; glycated he- moglobin reduction, -0.9 %; hypertension prevalence, 80 %) [10,28–34]. In this type of trials, oral antidiabetic agents were ad- ministered usually in addition to insulin. However, insulin was used more frequently in the intensified antidiabetic treatment arm and less frequently in those randomized to usual treatment. The systolic/dia- stolic BP reduction between arms was not significant (Fig. 1). In the two trials [30,32] with no or minimal representation of hypertensive pa- tients BP-lowering was also not evident.

11. Does BP-change following antidiabetic drug treatment explain cardiovascular risk reduction?

This question should be seen in the context of how BP-lowering by any agent is trailing back-and-forth specific outcome risk reduction. The effects of BP-lowering on fatal and non-fatal outcomes, as assessed by a comprehensive previous meta-regression analysis of anti- hypertensive drug treatment [55] is summarized in Fig. 2. For example, in the Canagliflozin Cardiovascular Assessment Study (CANVAS) Pro- gram [24], the observed 33 % (95 % confidence interval, 13 %–58 %) risk reduction in heart failure requiring hospitalization can be seen under the observation that the attained systolic/diastolic BP reduction during the trial was -3.9/-1.4 mmHg (i.e.−2.7mmHg is the averaged BP reduction between the systolic and diastolic BP). In Fig. 2, it is shown that the incidence of heart failure is reduced by 2.7 % for 1mmHg BP reduction. Thus, the BP-independent mean effect (i.e. at 0/ 0mmHg systolic/diastolic BP difference) of canagliflozin on heart

failure in the CANVAS-program trial would be almost 5% lower com- pared to the observed, that is -28 %. In Table 2, we provide the ex- pected outcome risk changes independently of BP-reduction observed in our primary analysis (see Fig. 1 and Supplementary Table S1) for the three drug classes (i.e. TZDs, GLP-1 agonists and SGLT2 inhibitors) in which a significant BP-lowering was noticed. It can be appreciated that BP-independent risk estimates were not significantly different com- pared with the initial analysis for TZDs and GLP1 agonists. Regarding SGLT2 inhibitors, risk reduction of 17 % in cardiovascular death faded away following the adjusted analysis.

12. Discussion

In this updated analysis we demonstrated that antidiabetic drugs in large outcome RCTs are accompanied by a subtle systolic BP reduction, while diastolic BP was not reduced. We also noticed that the higher extent of glucose lowering (more vs less intense trial design) was not associated with BP reduction. However, SGLT2-inhibitors, GLP-1 ago- nist and TZDs demonstrated a significant BP reduction in outcome RCTs, but the extent of BP lowering cannot be compared because is largely indirect. Only, head-to-head trials of these 3 classes of anti- diabetic agents can resolve the issue of whether one drug class may lower BP levels to a higher extent compared to another. Our findings should be also seen under the following considerations. First, the active or more intense glucose-lowering arm group of patients was followed by more trial visits and BP control was promptly adjusted compared with the usual glucose-lowering group in which follow-up visits were fewer and any BP control adjustment was delayed. The similar dis- tribution of baseline antihypertensive agents among randomized arms in glucose-lowering RCTs may be altered during follow-up, because additional agents or increased dosing of those already received may have been prescribed during follow-up visits. Consequently, the re- sulting BP difference cannot be exclusively ascribed to hemodynamic effects of antidiabetic agents, but also to different management of hy- pertension during follow-up, which may also contribute to a “by chance” BP-lowering. Second, the fact that a BP-lowering in T2DM patients without hypertension was indifferent, indicates that the po- tential beneficial effects of glucose-lowering on the vasculature are evident only in hypertensive states, in which vascular damage is more pronounced or a synergistic effect between antidiabetic and anti- hypertensive agents may play an important role in this setting. Third,

Fig. 2. Extent of outcome reduction with a systolic or diastolic blood pressure-lowering of 1 mmHg by any blood pressure-lowering drug.

I. Ilias, et al. Pharmacological Research 161 (2020) 105108

6

BP-measurements in glucose lowering outcome trials were performed with different and largely unknown methodology because BP-reduction was not a pre-specified outcome in most trials. Thus, any conclusion about BP-lowering at a single trial or pooled trial level should con- sidered with caution. Fourth, the observed BP-lowering effect in out- come trials is different from that reported in glucose control efficacy trials, because in the second case the calculated BP difference from baseline is related to baseline BP. Lastly, background antidiabetic agent regimen used in different glucose-lowering trials may have opposite effects on BP compared with the trial glucose-lowering drug and thus BP changes are may be reciprocally eliminated.

A raising hypothesis of the present overview is that BP-lowering of antidiabetic treatment cannot significantly contribute to cardiovascular benefits specifically observed with the new era antidiabetic drugs. It remains open which might be the pivotal mechanism to determine cardiovascular protection in high-risk T2DM patients already receiving optimal background antihypertensive, antiplatelet and antilipidemic treatment. The anti-atherosclerotic effects of the GLP-1 agonists may be seen as intermediate mediators of cardiovascular protection [8,51], but these mechanisms seem at least neutral following the use of SGLT-2 inhibitors [7]. Whether GLP1 agonists or SGLT2 inhibitors confer higher cardiovascular protection may be examined in future head-to- head RCTs. A direct comparison design can address which drug reduces in a higher extent systolic BP, since diastolic BP is not reduced by GLP1 agonists vs placebo, but can also unmask biologically plausible patho- physiological mechanisms linked to cardiovascular protection.

The finding that SGLT2 inhibitors showed a consistent BP-lowering across studies but did not prevent stroke which is the most elevated BP- dependent outcome compared to others, is difficult to explain. However, it might be hypothesized that BP-lowering in diabetes may be related to volume depletion and hypotensive phenomena [24] that in turn, may reduce cerebral perfusion and counterbalance any protective effect related to BP-lowering. In addition, the neutral effect of SGLT2 inhibitors on stroke should be interpreted in the context of two different lines of evidence retrieved from antihypertensive drug trials in- vestigating the comparison between more vs less BP-lowering targets [56]. In the Action to Control Cardiovascular Risk in Diabetes (AC- CORD) trial [57], a BP-lowering of -14.2/-6.4 mmHg was not associated with prevention of the primary composite cardiovascular outcome, whilst showed a significant reduction of strokes. However, in the AC- CORD trial [57], stroke was only a prespecified secondary outcome, while stroke event rate was rather limited. By contrast, in the Systolic Blood Pressure Intervention trial (SPRINT) [58], a BP reduction of

-13.1/-7.8 mmHg in patients without diabetes mellitus and history of cerebrovascular disease was not associated with stroke prevention, while hypotension rates were higher in the aggressive treatment arm. The indifferent effect of SGLT2 inhibitors on stroke prevention contrasts with the beneficial effect of GLP1 agonists on the same outcome but with a modest systolic BP reduction. More research is needed to resolve the above controversy and determine whether GPL1 agonists may exert specific cerebrovascular protection.

We acknowledge some limitations. In all selected RCTs, the mea- surement of BP was not clearly defined and the implemented metho- dology may vary across studies (outcome related bias). Most, of the RCTs considered here were industry sponsored and averaged outcome risk estimates may be inflated. In trials designed to evaluate more vs less aggressive blood glucose lowering the unblinded nature may have affected the results since the more active group was observed more intensively. Another limitation is that our meta-analysis, though sti- mulated by the need to answer questions relevant to medical practice, have put together trials that were not intended to investigate our principal research question, as unfortunately no trial approached the BP-lowering effect of antidiabetic drug classes. Finally, meta-analyses are hypothesis raising instruments and should not be seen as substitutes of well-designed RCTs.

13. Conclusion

Most of antidiabetic drugs including insulin have mild or neutral effects on BP. The exception to this general rule is the class of SGLT-2 inhibitors that can reduce systolic and diastolic BP by almost 4 and 1mmHg, respectively. SGLT2 inhibitors can be seen as a new class of diuretic agents. In most if not all cases, outcome risk reduction observed in glucose-lowering RCTs cannot be justified by a BP-lowering effect of antidiabetic drugs.

Declaration of Competing Interest

All authors declare no conflict of interest regarding the present work and they have no involvements that might raise the question of bias in the work reported or in the conclusions, implications or opinions stated.

Acknowledgements

None.

Table 2 Observed and expected (adjusted) risk ratios of various outcomes in randomized clinical trials of specific antidiabetic drug classes.

Outcomes Trials, n Observed SBP/DBP reduction, mmHg Observed RR (95 % CI) Adjusted RR (95 % CI) TZDs

CHD 0 NA – – Stroke 1 NA – – HF 2 −2.5/0 1.32 (1.12−1.54) 1.37 (1.16−1.60) CV Death 2 −2.5/0 1.03 (0.85−1.23) 1.05 (0.87−1.26) All-cause death 2 −2.5/0 1.01 (0.87−1.17) 1.02 (0.88−1.19)

GLP-1 agonists CHD 5 −1.2/0 0.92 (0.83−1.02) 0.93 (0.84−1.03) Stroke 5 −1.2/0 0.86 (0.78−0.95) 0.87 (0.79−0.96) HF 5 −1.2/0 0.92 (0.83−1.01) 0.93 (0.84−1.02) CV Death 6 −1.3/0 0.90 (0.83−0.97) 0.90 (0.83−0.98) All-cause death 6 −1.3/0 0.90 (0.85−0.95) 0.90 (0.85−0.95)

SGLT2 inhibitors CHD 3 −3.5/-1 0.89 (0.82−0.96) 0.92 (0.85−0.99) Stroke 3 −3.5/-1 0.97 (0.79−1.19) 1.03 (0.84−1.26) HF 4 −3.5/-1 0.67 (0.61−0.73) 0.71 (0.65−0.78) CV Death 4 −3.5/-1 0.83 (0.69−0.99) 0.86 (0.72−1.02) All-cause death 4 3.5/-1 0.84 (0.76−0.92) 0.86 (0.78−0.94)

CHD, coronary heart disease; CI, confidence interval; CV, cardiovascular; DBP, diastolic blood pressure; HF, heart failure; GLP1, glucagon-like peptide-1 NA, not available; n, number; RR, risk ratios; SBP, systolic blood pressure; SGLT2, sodium-glucose cotransporter 2.

I. Ilias, et al. Pharmacological Research 161 (2020) 105108

7

Appendix A. Supplementary data

Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.phrs.2020.105108.

References

[1] I.M. Stratton, C.A. Cull, A.I. Adler, D.R. Matthews, H.A. Neil, R.R. Holman, Additive effects of glycaemia and blood pressure exposure on risk of complications in type 2 diabetes: a prospective observational study (UKPDS 75), Diabetologia 49 (2006) 1761–1769.

[2] B. Williams, G. Mancia, W. Spiering, E.A. Rosei, M. Azizi, M. Burnier, et al., 2018 ESC/ESH guidelines for the management of arterial hypertension, Eur. Heart J. (39) (2018) 3021–3104.

[3] B. Hemmingsen, S.S. Lund, C. Gluud, A. Vaag, T.P. Almdal, J. Wetterslev, Targeting intensive glycaemic control versus targeting conventional glycaemic control for type 2 diabetes mellitus, Cochrane Database Syst. Rev. (2011) CD008143.

[4] R.R. Holman, S.K. Paul, M.A. Bethel, D.R. Matthews, H.A. Neil, 10-year follow-up of intensive glucose control in type 2 diabetes, N. Engl. J. Med. 359 (2008) 1577–1589.

[5] S.E. Nissen, K. Wolski, Effect of rosiglitazone on the risk of myocardial infarction and death from cardiovascular causes, N. Engl. J. Med. 356 (2007) 2457–2471.

[6] U.S. Food and Drug Administration, Guidance for Industry: Diabetes Mellitus Evaluating Cardiovascular Risk in New Antidiabetic Therapies to Treat Type 2 Diabetes, (2008) Available at:www.fda.gov/downloads/Drugs/ GuidanceComplianceRegulatoryInformation/Guidances/ucm071627.pdf Accessed April 24, 2020.

[7] A.J. Scheen, Cardiovascular effects of new oral glucose-lowering agents: DPP-4 and SGLT-2 inhibitors, Circ. Res. 122 (10) (2018) 1439–1459.

[8] J.D. Newman, A.K. Vani, J.O. Aleman, H.S. Weintraub, J.S. Berger, A.Z. Schwartzbard, The changing landscape of diabetes therapy for cardiovascular risk reduction: JACC state-of-the-Art review, J. Am. Coll. Cardiol. 72 (15) (2018) 1856–1869 October 9.

[9] H.C. Gerstein, J. Bosch, G.R. Dagenais, R. Dı´az, H. Jung, A.P. Maggioni, et al., Basal insulin and cardiovascular and other outcomes in dysglycemia, N. Engl. J. Med. 367 (2012) 319–328.

[10] A. Patel, S. MacMahon, J. Chalmers, B. Neal, L. Billot, M. Woodward, et al., Intensive blood glucose control and vascular outcomes in patients with type 2 diabetes, N. Engl. J. Med. 358 (2008) 2560–2572.

[11] A. Kooy, J. de Jager, P. Lehert, D. Bets, M.G. Wulffele´, A.J. Donker, C.D. Stehouwer, Long-term effects of metformin on metabolism and microvascular and macrovascular disease in patients with type 2 diabetes mellitus, Arch. Intern. Med. 169 (2009) 616–625.

[12] J.A. Dormandy, B. Charbonnel, D.J. Eckland, E. Erdmann, M. Massi-Benedetti, I.K. Moules, et al., Secondary prevention of macrovascular events in patients with type 2 diabetes in the PROactive Study (PROspective pioglitAzone Clinical Trial in macrovascular Events): a randomised controlled trial, Lancet 366 (2005) 1279–1289.

[13] A.M. Lincoff, J.C. Tardif, G.G. Schwartz, S.J. Nicholls, L. Ryde´n, B. Neal, et al., Effect of aleglitazar on cardiovascular outcomes after acute coronary syndrome in patients with type 2 diabetes mellitus: the AleCardio randomized clinical trial, JAMA 311 (2014) 1515–1525.

[14] J. Rosenstock, V. Perkovic, O.E. Johansen, M.E. Cooper, S.E. Kahn, N. Marx, et al., Effect of linagliptin vs placebo on major cardiovascular events in adults with type 2 diabetes and high cardiovascular and renal risk: the CARMELINA randomized clinical trial, JAMA 321 (2019) 69–79.

[15] W.B. White, C.P. Cannon, S.R. Heller, S.E. Nissen, R.M. Bergenstal, G.L. Bakris, et al., Alogliptin after acute coronary syndrome in patients with type 2 diabetes, N. Engl. J. Med. 369 (2013) 1327–1335.

[16] B.M. Scirica, D.L. Bhatt, E. Braunwald, P.G. Steg, J. Davidson, B. Hirshberg, et al., Saxagliptin and cardiovascular outcomes in patients with type 2 diabetes mellitus, N. Engl. J. Med. 369 (2013) 1317–1326.

[17] J.B. Green, M.A. Bethel, P.W. Armstrong, J.B. Buse, S.S. Engel, J. Garg, et al., Effect of sitagliptin on cardiovascular outcomes in type 2 diabetes, N. Engl. J. Med. 373 (2015) 232–242.

[18] M.A. Pfeffer, B. Claggett, R. Diaz, K. Dickstein, H.C. Gerstein, L.V. Køber, et al., Lixisenatide in patients with type 2 diabetes and acute coronary syndrome, N. Engl. J. Med. 373 (2015) 2247–2257.

[19] R.R. Holman, M.A. Bethel, R.J. Mentz, V.P. Thompson, Y. Lokhnygina, J.B. Buse, et al., Effects of once-weekly exenatide on cardiovascular outcomes in type 2 dia- betes, N. Engl. J. Med. 377 (2017) 1228–1239.

[20] S.P. Marso, G.H. Daniels, K. Brown-Frandsen, P. Kristensen, J.F. Mann, M.A. Nauck, et al., Liraglutide and cardiovascular outcomes in type 2 diabetes, N. Engl. J. Med. 375 (2016) 311–322.

[21] A.F. Hernandez, J.B. Green, S. Janmohamed, R.B. D’Agostino Sr, C.B. Granger, N.P. Jones, et al., Albiglutide and cardiovascular outcomes in patientswith type 2 diabetes and cardiovascular disease (Harmony Outcomes): a double-blind, rando- mized placebo-controlled trial, Lancet 392 (2018) 1519–1529.

[22] H.C. Gerstein, H.M. Colhoun, G.R. Dagenais, R. Diaz, M. Lakshmanan, P. Pais, J. Probstfield, F.T. Botros, M.C. Riddle, L. Rydén, D. Xavier, C.M. Atisso, L. Dyal, S. Hall, P. Rao-Melacini, G. Wong, A. Avezum, J. Basile, N. Chung, I. Conget, W.C. Cushman, E. Franek, N. Hancu, M. Hanefeld, S. Holt, P. Jansky, M. Keltai, F. Lanas, L.A. Leiter, P. Lopez-Jaramillo, E.G. Cardona Munoz, V. Pirags, N. Pogosova, P.J. Raubenheimer, J.E. Shaw, W.H. Sheu, T. Temelkova-Kurktschiev,

REWIND Investigators, Dulaglutide and renal outcomes in type 2 diabetes: an ex- ploratory analysis of the REWIND randomised, placebo-controlled trial, Lancet 394 (10193) (2019) 131–138 July 13.

[23] S.P. Marso, S.C. Bain, A. Consoli, F.G. Eliaschewitz, E. Jo´dar, L.A. Leiter, et al., Semaglutide and cardiovascular outcomes in patients with type 2 diabetes, N. Engl. J. Med. 375 (2016) 1834–1844.

[24] B. Neal, V. Perkovic, K.W. Mahaffey, D. de Zeeuw, G. Fulcher, N. Erondu, et al., Canagliflozin and cardiovascular and renal events in type 2 diabetes, N. Engl. J. Med. 377 (2017) 644–657.

[25] S.D. Wiviott, I. Raz, M.P. Bonaca, O. Mosenzon, E.T. Kato, A. Cahn, et al., Dapagliflozin and cardiovascular outcomes in type 2 diabetes, N. Engl. J. Med. 380 (2019) 347–357.

[26] B. Zinman, C. Wanner, J.M. Lachin, D. Fitchett, E. Bluhmki, S. Hantel, et al., Empagliflozin, cardiovascular outcomes, and mortality in type 2 diabetes, N. Engl. J. Med. 373 (2015) 2117–2128.

[27] V. Perkovic, M.J. Jardine, B. Neal, et al., Canagliflozin and renal outcomes in type 2 diabetes and nephropathy, N. Engl. J. Med. 380 (24) (2019) 2295–2306.

[28] H.C. Gerstein, M.E. Miller, R.P. Byington, D.C. Goff Jr, J.T. Bigger, J.B. Buse, et al., Effects of intensive glucose lowering in type 2 diabetes, N. Engl. J. Med. 358 (2008) 2545–2559.

[29] R.L. Frye, P. August, M.M. Brooks, R.M. Hardison, S.F. Kelsey, J.M. MacGregor, et al., A randomized trial of therapies for type 2 diabetes and coronary artery dis- ease, N. Engl. J. Med. 360 (2009) 2503–2515.

[30] M. Shichiri, H. Kishikawa, Y. Ohkubo, N. Wake, Long-term results of the Kumamoto study on optimal diabetes control in type 2 diabetic patients, Diabetes Care 23 (Suppl 2) (2000) B21–B29.

[31] J.C. Chan, W.Y. So, C.Y. Yeung, G.T. Ko, I.T. Lau, M.W. Tsang, et al., Effects of structured versus usual care on renal endpoint in type 2 diabetes: the SURE study: a randomized multicenter translational study, Diabetes Care 32 (2009) 977–982.

[32] UK Prospective Diabetes Study (UKPDS) Group, Intensive blood glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33), Lancet 352 (1998) 837–853.

[33] C. Abraira, W.G. Henderson, J.A. Colwell, F.Q. Nuttall, J.P. Comstock, N.V. Emanuele, et al., Response to intensive therapy steps and to glipizide dose in combination with insulin in type 2 diabetes. VA feasibility study on glycemic control and complications (VA CSDM), Diabetes Care 21 (1998) 574–579.

[34] W. Duckworth, C. Abraira, T. Moritz, D. Reda, N. Emanuele, P.D. Reaven, et al., Glucose control and vascular complications in veterans with type 2 diabetes, N. Engl. J. Med. 360 (2009) 129–139.

[35] C. Thomopoulos, G. Bazoukis, I. Ilias, C. Tsioufis, T. Makris, Effects of glucose- lowering on outcome incidence in diabetes mellitus and the modulating role of blood pressure and other clinical variables: overview, meta-analysis of randomized trials, J. Hypertens. 37 (October 10) (2019) 1939–1949.

[36] W.T. Cefalu, S. Kaul, H.C. Gerstein, R.R. Holman, B. Zinman, J.S. Skyler, et al., Cardiovascular outcomes trials in type 2 diabetes: where do we go from here? Reflections from a diabetes care editors’ expert forum, Diabetes Care 41 (2018) 14–31.

[37] L.R. Kurukulasuriya, S. Stas, G. Lastra, C. Manrique, J.R. Sowers, Hypertension in obesity, Med. Clin. North Am. 95 (2011) 903–917.

[38] E. Ferrannini, W.C. Cushman, Diabetes and hypertension: the bad companions, Lancet 380 (9841) (2012) 601–610. August 11.

[39] A.M. Heagerty, E.H. Heerkens, A.S. Izzard, Small artery structure and function in hypertension, J. Cell. Mol. Med. 14 (2010) 1037–1043.

[40] H.A. Randeree, M.A. Omar, A.A. Motala, M.A. Seedat, Effect of insulin therapy on blood pressure in NIDDM patients with secondary failure, Diabetes Care 15 (October 10) (1992) 1258–1263.

[41] T. Heise, K. Magnusson, L. Heinemann, P.T. Sawicki, Insulin resistance and the effect of insulin on blood pressure in essential hypertension, Hypertension 32 (August 2) (1998) 243–248.

[42] A.K. Singh, R. Singh, Is gliclazide a sulfonylurea with difference? A review in 2016, Expert Rev. Clin. Pharmacol. 9 (June 6) (2016) 839–851.

[43] J. Rosenstock, S.E. Kahn, O.E. Johansen, B. Zinman, M.A. Espeland, H.J. Woerle, E. Pfarr, A. Keller, M. Mattheus, D. Baanstra, T. Meinicke, J.T. George, M. von Eynatten, D.K. McGuire, N. Marx, CAROLINA Investigators, Effect of Linagliptin vs Glimepiride on major adverse cardiovascular outcomes in patients with type 2 diabetes: The CAROLINA randomized clinical trial, JAMA (September 19) (2019), https://doi.org/10.1001/jama.2019.13772.

[44] C. Thomopoulos, G. Katsimagklis, T. Makris, Metformin and blood pressure low- ering: a questioned association, J. Hypertens. 35 (January 1) (2017) 27–28.

[45] L. Zhou, H. Liu, X. Wen, Y. Peng, Y. Tian, L. Zhao, Effects of metformin on blood pressure in nondiabetic patients: a meta-analysis of randomized controlled trials, J. Hypertens. 35 (January 1) (2017) 18–26.

[46] M.G. Wulffele´, A. Kooy, D. de Zeeuw, C.D. Stehouwer, R.T. Gansevoort, The effect of metformin on blood pressure, plasma cholesterol and triglycerides in type 2 diabetes mellitus: a systematic review, J. Intern. Med. 256 (2004) 1–14.

[47] T. Roszer, M. Ricote, PPARs in the renal regulation of systemic blood pressure, PPAR Res. 2010 (2010) 698730.

[48] W.J. Geese, W. Achanzar, C. Rubin, N. Hariharan, P. Cheng, L. Tomlinson, N. Ordway, N.C. Dracopoli, T. Delmonte, L. Hui, B. Krishnan, G. Cosma, K. Ranade, Genetic and gene expression studies implicate renin and endothelin-1 in edema caused by peroxisome proliferator-activated receptor gamma agonists, Pharmacogenet. Genomics 18 (October 10) (2008) 903–910.

[49] M. Packer, Worsening heart failure during the use of DPP-4 inhibitors: pathophy- siological mechanisms, clinical risks, and potential influence of concomitant anti- diabetic medications, JACC Heart Fail 6 (June 6) (2018) 445–451.

I. Ilias, et al. Pharmacological Research 161 (2020) 105108

8

[50] X. Zhang, Q. Zhao, Effects of dipeptidyl peptidase-4 inhibitors on blood pressure in patients with type 2 diabetes: a systematic review and meta-analysis, J. Hypertens. 34 (February 2) (2016) 167–175.

[51] A. Goud, J. Zhong, M. Peters, R.D. Brook, S. Rajagopalan, GLP-1 agonists and blood pressure: a review of the evidence, Curr. Hypertens. Rep. 18 (February 2) (2016) 16.

[52] M. Katout, H. Zhu, J. Rutsky, P. Shah, R.D. Brook, J. Zhong, S. Rajagopalan, Effect of GLP-1 mimetics on blood pressure and relationship to weight loss and glycemia lowering: results of a systematic meta-analysis and meta-regression, Am. J. Hypertens. 27 (January 1) (2014) 130–139.

[53] W.L. Baker, L.R. Smyth, D.M. Riche, E. Bourret, K.W. Chamberlin, W.B. White, Effects of sodium-glucose co-transporter 2 inhibitors on blood pressure: a systematic review and meta-analysis, J. Am. Soc. Hypertens. 8 (April 4) (2014) 262–275 e9.

[54] T.A. Zelniker, E. Braunwald, Cardiac and renal effects of sodium-glucose Co- transporter 2 inhibitors in diabetes: JACC state-of-the-Art review, J. Am. Coll.

Cardiol. 72 (15) (2018) 1845–1855 October 9. [55] C. Thomopoulos, G. Parati, A. Zanchetti, Effects of blood pressure lowering on

outcome incidence in hypertension. 1. Overview, meta-analyses, and meta-regres- sion analyses of randomized trials, J. Hypertens. 32 (December 12) (2014) 2285–2295.

[56] C. Thomopoulos, G. Parati, A. Zanchetti, Effects of blood pressure lowering on outcome incidence in hypertension: 7. Effects of more vs. less intensive blood pressure lowering and different achieved blood pressure levels - updated overview and meta-analyses of randomized trials, J. Hypertens. 34 (April 4) (2016) 613–622.

[57] The ACCORD Study Group, Effects of intensive blood-pressure control in type 2 diabetes mellitus, N. Engl. J. Med. 362 (2010) 1575–1585.

[58] J.T. Wright Jr, J.D. Williamson, P.K. Whelton, J.K. Snyder, K.M. Sink, M.V. Rocco, et al., A randomized trial of intensive versus standard blood-pressure control, N. Engl. J. Med. 373 (2015) 2103–2116.

I. Ilias, et al. Pharmacological Research 161 (2020) 105108

9