Antihypertensive Drugs in Arterial 'De-stiffening'
Antihypertensive Drugs in Arterial 'De-stiffening'
Several pharmacologic studies have evaluated the effects of different antihypertensive drug classes on arterial stiffness using variations in methodology (mainly in assessing and recording CBP, treatment duration, study design, and drug doses). As modifications in arterial wall properties occur after long-term drug treatment, there are considerable limitations regarding the extrapolation of available data to daily clinical practice. Therefore, it is not possible to draw definite conclusions for all classes of antihypertensive drugs. Nevertheless, the presence of consistently reproduced results can lead to significant comments on class effects.
Intra-arterial infusion of captopril in hypertensive patients caused a significant improvement in large artery compliance, leading to augmented carotid artery distensibility and reduced wave reflections. The administration of quinapril to subjects with essential hypertension showed that changes regarding carotid stiffness were exclusively due to BP reduction and not to a drug-induced relaxation of the arterial wall, whereas alterations in aortic distensibility were attributable to both factors, suggesting that this effect was partly independent of BP reduction.
Nevertheless, it was not until the REASON (pREterax in regression of Arterial Stiffness in a contrOlled double-bliNd) study that a large clinical trial focused attention on the longterm effects of antihypertensive drugs on pulsatile hemodynamics in essential hypertension. The combination of perindopril with a low dose of indapamide (PER/IND) was compared with atenolol over 12 months of treatment. It was shown that, for a similar fall of MAP between the two-arm groups, a significantly greater reduction of SBP and PP was obtained with PER/IND compared with atenolol. Furthermore, the effect on central indices was more pronounced than in the peripheral indices. Despite the fact that PWV was equally reduced by the two drug regimens, only PER/IND was associated with a significant reduction in AIx. In addition, subjects in the arm of the PER/IND-based regimen had substantially decreased cardiac hypertrophy, and this was strongly correlated to reduced central wave reflections, whereas in the atenolol arm this effect was less pronounced. Interestingly, the remodelling effect of the combination regimen on arterial wall properties was sustained even after 9 months of treatment, without any further reduction of PBP.
PER/IND was also evaluated in the ADVANCE (Action in Diabetes and Vascular Disease: Preterax and Diamicron Modified Release Controlled Evaluation) trial. The outcomes, in agreement with those of the REASON study, strengthened the hypothesis that SBP and PP are strong predictors of CV events in diabetic patients. Moreover, baseline arterial stiffness was shown to be the most important predictor of CV risk.
The CAFE (Conduit Artery Functional Evaluation) study, a subanalysis of the controlled ASCOT (Anglo-Scandinavian Cardiac Outcomes Trial) in which 2073 subjects were randomized to receive either atenolol adding bendrofluazide as required or amlodipine adding perindopril as required, showed that aortic (central) PP, recorded non-invasively by radial tonometry and the application of generalized transfer functions, is a major determinant of clinical outcomes independent of age, other traditional CV risk factors, and even peripheral PP. In line with the REASON trial, the CAFE study results reported a different impact on CBP and indices, which were reported as significantly lower in the amlodipine-based regimen, despite the identical reduction of SBP and PP produced by the two regimens. Furthermore, this differencewas sustained over the 4-year follow-up.
The effect of ACE inhibition on pulsatile hemodynamics in patients with stable coronary disease and preserved left ventricular function was assessed in the PEACE (Prevention of Events with Angiotensin Converting Enzyme inhibition) substudy, which evaluated the response to trandolapril administration, regarding its effect on CBP and arterial stiffness. The carotid-femoral PWV was moderately decreased, but beyond expectations according to BP reduction or differences in baseline characteristics alone. Of note, no modification in aortic compliance, Aix, or total arterial compliance was reported.
The outcomes of the vascular substudy of the ANBP2 (second Australian National Blood Pressure Study) in an elderly hypertensive population had been much anticipated, due to the fact that, for the first time, the recording of CBP was performed before randomization in a large clinical trial. A subset (479 patients) of the trial cohort was evaluated both before and after 4 years of treatment. Although ANBP2 reported better prognosis for male (no beneficial effect was evident in women) elderly (65–84 years old) hypertensive subjects randomly assigned to the ACEI enalapril compared with a diuretic-based regimen, no differential effect on CBP and PBP was observed between the two arms. These results may be partly explained by the increased arterial stiffness, which occurs with age and hypertension.
In addition to the aforementioned long-term important studies, several others, most of them mid- (less than 6 months' duration) or short- (less than 4 weeks' duration, including acute effect studies lasting few hours) term, evaluated the effect of ACEIs (either as monotherapy, or as a single add-on treatment, or even head-to-head comparison). These studies provided convincing evidence that the reduction observed in arterial stiffness (as measured by PP, PWV, and AIx) and the positive effect on BP amplification should be considered a class effect (Table I). Furthermore, evidence clearly supports the presence of an additional CBP-lowering ability by ACEIs, beyond PBP, although this effect is not present in all the available studies. The authors proposed that the differential effect ofACEIs onCBP may be explained on the basis of their binding affinity for tissue ACE (e.g. quinapril has higher affinity than captopril) and their effect on endothelial function. Despite their minor differences, ACEIs seem to act by decreasing wave reflections potentially attributable to chronic inverse remodelling of the small arteries leading to reduced reflection coefficients.
The aforementioned data regarding the effects of ACE inhibition on pulsatile hemodynamics suggest that ARBs, affecting the same pressor system, might produce similar results (Table II). In patients with resistant hypertension (receiving at least three antihypertensive drugs at maximum dosage, including an ACEI), the addition of valsartan for 2 weeks significantly reduced AIx. Arandomized, crossover study in hypertensive patients compared valsartan with captopril. The two drugs equally reducedPWVand AIx. In addition, their combination resulted in further reductions in arterial stiffness. Most importantly, this effect remained significant even after BP adjustment, implying that different drugs that independently improve arterial wall properties may lead to a greater reduction in arterial stiffness when combined.
In a 4-week crossover study, losartan was compared with hydrochlorothiazide (HCTZ) in patients with essential hypertension. The authors showed that only losartan reduced PWV and aortic AIx and augmented PP amplification, for comparable BP reduction. Similar results were reported for valsartan. Telmisartan was evaluated in hypertensive patients with type 2 diabetes mellitus, and was shown to substantially reduce PWV and AIx compared with placebo.
In hypertensive patients, eprosartan significantly decreased both PBP and CBP compared with baseline, as evidenced by the PP amplification ratio (peripheral/central PP ratio: 1.38 vs 1.41, p < 0.005), which was consistent with a CBP reduction beyond the PBP reduction. Of note, a decrease in absolute PP amplification was reported after eprosartan administration (15mmHg vs 13mmHg at baseline), creating reasonable doubt regarding the efficacy of eprosartan in reducing CBP beyond PBP. Nevertheless, wave reflections (AIx) and aortic stiffness (PWV) were significantly reduced. Olmesartan produced similar results in a different clinical trial when administered to patients with mild to moderate essential hypertension.
In another study, 15 hypertensive patients were given irbesartan or lisinopril for 12 weeks and then crossed over for 12 weeks. Arterial stiffness was assessed by measuring PWV in the carotid-femoral (CF), carotid-radial (CR), and femoral dorsalis-pedis (FD). While irbesartan and lisinopril reduced PWV(CF) in the elastic arterial system, only irbesartan reduced PWV (CR) and PWV (FD), enhancing arterial compliance in both elastic and muscular arteries. The difference between treatments remained significant after SBP adjustment.
Whereas the direct association between arterial stiffness reduction and decreased CV risk has been efficiently documented with ACEIs, it remains unclear with ARBs. However, important equivalence studies support the potential hypothesis that ARBs may reduce primary CV outcomes due to beneficial effects that are beyond PBP lowering.
In the LIFE (Losartan Intervention For Endpoint reduction in hypertension) study, losartan was shown to be superior to atenolol in preventing CV outcomes, but the effect was more pronounced than expected from the decrease in PBP. It has been speculated that this protective effect might be attributable to direct additional properties of losartan on arterial wall structure. Furthermore, in a small randomized, doubleblind, crossover study with a similar population to LIFE, it was shown that BNP was significantly lower in the losartan arm compared with atenolol after a 4-month treatment, indicating that the survival benefit reported for losartan in the LIFE study could be attributed to the observed improvement of BNP levels. Nevertheless, CBP and indices were not recorded and the authors reported no significant difference between losartan and atenolol in terms of AIx and PWV.
The VALUE (Valsartan Antihypertensive Long-Term Use Evaluation) study, which compared the potential effects of a valsartan-based regimen to an amlodipine-based regimen with regard to their effects on cardiac morbidity and mortality in hypertensive individuals at high risk of cardiac events, reported no difference in cardiac endpoints between the two treatment groups. However, in a subanalysis of the VALUE study, in which patients had to receive monotherapy with either of the two study drugs, despite similar BP reductions in the two subgroups, relative risks of heart failure and new-onset diabetes were in favor of the valsartan-based treatment. Finally, in the OPTIMAAL (Optimal Trial In Myocardial Infarction with the Angiotensin Antagonist Losartan) study, losartan was shown to share with captopril equally beneficial effects in reducing all-cause mortality in high-risk patients after myocardial infarction. Of note, pulsatile hemodynamics (i.e. CBP and indices, AIx and PWV) were not assessed in the aforementioned equivalence studies; consequently, no direct linking may be presumed between their results and possible modifications in arterial wall properties.
Pleiotropic effects have been attributed to both drug classes. Thus, it seems relevant to propose that ARBs may have beneficial effects on CBP beyond PBP lowering. The available data on their potential long-term effect on arterial stiffness is still weak and further clinical evidence is required.
In a randomized, 1-month study in a hypertensive population comparing spironolactone to bendrofluazide, only spironolactone decreased PWV and AIx and these effects remained significant after adjustment for BP change. Another randomized, double-blind trial evaluated the administration of either spironolactone or HCTZ as monotherapy in older hypertensive patients for 6 months, with a significant reduction in BP, PP, and PWV, which was comparable in both groups. Spironolactone was also shown to improve arterial stiffness in early-stage chronic kidney disease.
In a direct head-to-head comparison, eplerenone was proven to share with amlodipine equally beneficial effects on SBP, PP, and PWV in an elderly population suffering from isolated systolic hypertension. For comparable BP reduction, eplerenone was associated with decreased vessel stiffness and collagen/elastin ratio and reduced circulating inflammatory markers, contrary to atenolol administration that was associated with an increase in resistant artery stiffness.
The hemodynamic effects of the first direct renin inhibitor (DRI), aliskiren, were only recently investigated. Arterial stiffness was assessed in adults with uncomplicated type 1 diabetes mellitus, via recording AIx and PWV before and after aliskiren administration, during clamped euglycemia and hyperglycemia. The carotid AIx during clamped euglycemia was shown to decrease in parallel to PWV during clamped hyperglycemia, suggesting that DRIs improve arterial wall properties. Nevertheless, further clinical research is required.
The CAFE study, which was discussed in section 5.1, provided the most important evidence on the effect of dihydropyridine CCBs (amlodipine) on pulsatile hemodynamics. Despite the lack of baseline data regarding CBP and central indices, CAFE is singled out as the only available large study relating CBP modifications due to antihypertensive treatment to mortality.
A prospective study evaluated the effects of a CCB (azelnidipine) versus a diuretic (HCTZ), both in combination with the same ARB (olmesartan). Patients who were administered olmesartan for 3 months were randomized to receive either azelnidipine or HCTZ for another 6 months. The authors found a significantly greater reduction in central SBP in the olmesartan/azelnidipine group, despite the fact that brachial SBP was equally reduced in both groups. The decrease in AIx was similar between the two groups; however, when adjusted for a heart rate (HR) of 75 beats/min, the reduction was significantly greater in the olmesartan/azelnidipine group. PP amplification in the olmesartan/azelnidipine group was significantly increased as well. Furthermore, a more significant decrease in aortic stiffness, expressed by the aortic PWV, was reported in the olmesartanazelnidipine arm. These results suggest a superiority of azelnidipine over HCTZ, when combined with the same dose of olmesartan, regarding their effect on central hemodynamics, although no difference between peripheral SBP reductions was reported.
Several studies have evaluated the effect of dihydropyridine CCBs on arterial stiffness (Table III). Nitrendipine was shown to significantly reduce both PBP and CBP in patients with endstage renal disease. However, the substantially enhanced PP amplification suggests a more pronounced effect on central PP. Nitrendipine was also associated with reduced PWV and AIx.
CBP and AIx were reduced when patients with newly diagnosed, mild to moderate hypertension were administered barnidipine for a 6-month period. However, no substantial changes in intrinsic arterial stiffness were observed.
In another study in untreated hypertensive patients, felodipine and amlodipine were shown to reduce CBP to a similar extent. Even though both peripheral and central SBP were reduced, the authors reported a more pronounced effect on central rather than peripheral SBP, compared with placebo, as well as on PP. Of note, the aforementioned effects were shown to be dose-dependent, since they were enhanced with a 10mg dosage (compared with 5 mg) of either drug. The wave reflections were also significantly reduced (−10%).
Relatively few studies have directly compared the four major antihypertensive drug classes (ACEIs, CCBs, BBs, and diuretics) on PBP, CBP, and arterial stiffness. In one randomized study, 59 elderly patients suffering from untreated isolated systolic hypertension were assigned to one of the following antihypertensive drugs: perindopril, atenolol, lercanidipine, or bendrofluazide. Central and peripheral BP, pressure wave reflections and aortic stiffness were assessed at baseline, following 2 weeks of placebo therapy, and at the end of 10 weeks of active therapy. Peripheral SBP and peripheral PP were reduced to a similar extent by all four drug classes. Furthermore, all but atenolol reduced central PP. Lercanidipine was the only agent that significantly reduced AIx, whereas atenolol increased it, though this effect was not significant after adjustment for HR. Aortic PWV remained unchanged by all four drugs.
Apart from dihydropyridine CCBs, verapamil was shown to reduce CBP, PP, and PWV in hypertensive patients.
Several of the aforementioned studies compared the effects of CCBs with other antihypertensive drugs on CBP and PP amplification, suggesting that CCBs are superior to BBs and diuretics, but inferior to ACEIs.
Despite the fact that diuretics are reported to increase arterial compliance, up to date data suggest that this effect is mainly attributed to the decline in BP rather than to a change in the intrinsic properties of the arterial wall.
The vascular substudy of the ANBP2 trial provided noteworthy evidence regarding CBP modifications in response to administration of diuretics. Sixty-four percent of the participants were treated with HCTZ (the final choice belonged to the family practitioner). At the conclusion of the 4-year followup period, both PBP and CBP were significantly decreased in the HCTZ arm; however, the effect was more pronounced peripherally than centrally (SBP −17/−15mmHg, PP −9/−6mmHg, respectively).
The effect of standard (25 mg) and high-dose (50 mg) HCTZ on pulsatile hemodynamics was evaluated in elderly untreated patients with essential hypertension. Both doses reduced PBP and CBP as well as PP to a similar extent compared with placebo. Consequently, SBP and PP amplification remained unchanged at the end of the study, suggesting that HCTZ affects central and peripheral pressure parameters equally.
Bendrofluazide significantly reduced PBP compared with placebo, but had no impact on CBP. Comparison of spironolactone with bendrofluazide showed that only spironolactone reduced PWV and AIx and these effects remained significant even after adjustment for PBP change. Both indapamide and bendrofluazide altered hemodynamic parameters, reducing MAP and AIx to a similar extent in patients with stroke. Indapamide was also evaluated in subjects with essential hypertension for 2 months and no difference was observed regarding the effect on PBP, CBP, and AIx.
The above-mentioned studies reveal a neutral rather than a favorable effect of diuretics on pulsatile hemodynamics (Table IV). Consequently, diuretics are now considered to be inferior to the newer antihypertensive agents (ACEIs, CCBs) that have an additional effect on CBP beyond PBP, enhancing BP amplification. To date there are no available data from drug comparison trials involving diuretics and ARBs. In contrast, diuretics, due to their neutral effect on CBP, seem to outclass certain BBs (i.e. atenolol and bisoprolol), which were found to reduce BP amplification.
Most of the available studies regarding the effect of BBs on pulsatile hemodynamics have used the cardioselective agent atenolol. The REASON study was the first clinical trial to provide clear proof that atenolol, despite reducing both CBP and PBP, had a more pronounced peripheral than central effect. Interestingly, central PP was actually slightly increased, although peripheral PP was substantially reduced. Of note, PWV was decreased by almost 1 msec, whereas AIx was increased by approximately 2.5%, which might be related to the HR deceleration by 8 beats/min. The aforementioned findings were consistent with a number of smaller short- to midterm studies, confirming the hypothesis that atenolol negatively affects both SBP and PP amplification and wave reflections, resulting in an increase in AIx. Instead, aortic stiffness, assessed by PWV, was significantly reduced.
The CAFE study, in agreement with the REASON study, provided sufficient evidence of the inferiority of atenolol to amlodipine, also relating the negative effects of atenolol on CBP and indices to adverse CV outcomes. Furthermore, a more pronounced effect was observed on PBP than CBP. Based on these results, meta-analyses have challenged the recommendation for classic BBs (namely atenolol and bisoprolol) as an optimal treatment for uncomplicated hypertension.
Whether such observations should be considered a class effect remains unclear. BBs represent a heterogeneous antihypertensive drug class (Table V). There is now accumulating evidence that newer BBs with peripheral vasodilator properties, such as nebivolol and carvedilol, may be more effective in improving central pulsatile measurements. In animalmodels, nebivolol, but not atenolol, was shown to improve arterial distensibility. In another study, nebivolol, in contrast to atenolol, increased PP amplification and reduced AIx, possibly due to increased NO levels that affect the small resistance arteries. On the contrary, another study showed that although atenolol increased central PP and AIx more than nebivolol, both drugs reduced PP amplification. When nebivolol was compared with enalapril in hypertensive diabetic patients, both agents were shown to possess beneficial properties regarding arterial compliance and glucose metabolism. In patients suffering from coronary artery disease, endothelial function was improved by nebivolol, though not by atenolol, following therapy for a month. In a recent, retrospective study, three groups of hypertensive patients treated with ARBs, carvedilol/nebivolol, or atenolol were compared. For similar PBP and PWV reduction, treatment with either vasodilator BBs or ARBs was associated with lower CBP and AIx compared with atenolol. Furthermore, atenolol significantly increased peripheral AIx compared with metoprolol, for comparable MAP reduction. Metoprolol showed non-inferiority to valsartan with regard to their effect on PWV in mildly hypertensive patients.
Further large-scale, long-term outcome studies are required to determine whether BBs, classic and/or newer, are associated with adverse clinical outcomes. Finally, although precautions exist against using classic BBs in hypertensive patients with multiple metabolic risk factors, such recommendations do not apply to the newer vasodilator BBs, at least to date.
Although nitrates were never considered as first-line antihypertensive treatment for essential hypertension, invasive studies assessing their acute effect on pulsatile hemodynamics have provided noteworthy data. In this context, two invasive studies reported a greater reduction in CBP compared with PBP after 5 minutes of low-dose nitrates, nitroglycerine (sublingually), or nitroprusside (intravenously). An increase in SBP amplification was observed in both studies. By contrast, AIx was significantly decreased.
In patients with isolated systolic hypertension, SBP and PP amplification were increased and AIx was substantially reduced after the administration of isosorbide mononitrate. Equally beneficial results were also reported in a small (n = 6) group of patients under long-term therapy with either captopril or eprosartan, suggesting that an alternative mechanism, other than RAAS inhibition, might be involved, potentially through the direct relaxation of the muscular conduit arteries. On the contrary, their effect on PWV was apparently inadequate or, at least, equivocal, implying that large arterial stiffness is minimally affected by nitrates
5. Effect of Antihypertensive Agents on Arterial Stiffness and Central Pulsatile Hemodynamics
Several pharmacologic studies have evaluated the effects of different antihypertensive drug classes on arterial stiffness using variations in methodology (mainly in assessing and recording CBP, treatment duration, study design, and drug doses). As modifications in arterial wall properties occur after long-term drug treatment, there are considerable limitations regarding the extrapolation of available data to daily clinical practice. Therefore, it is not possible to draw definite conclusions for all classes of antihypertensive drugs. Nevertheless, the presence of consistently reproduced results can lead to significant comments on class effects.
5.1 Angiotension-converting Enzyme Inhibitors
Intra-arterial infusion of captopril in hypertensive patients caused a significant improvement in large artery compliance, leading to augmented carotid artery distensibility and reduced wave reflections. The administration of quinapril to subjects with essential hypertension showed that changes regarding carotid stiffness were exclusively due to BP reduction and not to a drug-induced relaxation of the arterial wall, whereas alterations in aortic distensibility were attributable to both factors, suggesting that this effect was partly independent of BP reduction.
Nevertheless, it was not until the REASON (pREterax in regression of Arterial Stiffness in a contrOlled double-bliNd) study that a large clinical trial focused attention on the longterm effects of antihypertensive drugs on pulsatile hemodynamics in essential hypertension. The combination of perindopril with a low dose of indapamide (PER/IND) was compared with atenolol over 12 months of treatment. It was shown that, for a similar fall of MAP between the two-arm groups, a significantly greater reduction of SBP and PP was obtained with PER/IND compared with atenolol. Furthermore, the effect on central indices was more pronounced than in the peripheral indices. Despite the fact that PWV was equally reduced by the two drug regimens, only PER/IND was associated with a significant reduction in AIx. In addition, subjects in the arm of the PER/IND-based regimen had substantially decreased cardiac hypertrophy, and this was strongly correlated to reduced central wave reflections, whereas in the atenolol arm this effect was less pronounced. Interestingly, the remodelling effect of the combination regimen on arterial wall properties was sustained even after 9 months of treatment, without any further reduction of PBP.
PER/IND was also evaluated in the ADVANCE (Action in Diabetes and Vascular Disease: Preterax and Diamicron Modified Release Controlled Evaluation) trial. The outcomes, in agreement with those of the REASON study, strengthened the hypothesis that SBP and PP are strong predictors of CV events in diabetic patients. Moreover, baseline arterial stiffness was shown to be the most important predictor of CV risk.
The CAFE (Conduit Artery Functional Evaluation) study, a subanalysis of the controlled ASCOT (Anglo-Scandinavian Cardiac Outcomes Trial) in which 2073 subjects were randomized to receive either atenolol adding bendrofluazide as required or amlodipine adding perindopril as required, showed that aortic (central) PP, recorded non-invasively by radial tonometry and the application of generalized transfer functions, is a major determinant of clinical outcomes independent of age, other traditional CV risk factors, and even peripheral PP. In line with the REASON trial, the CAFE study results reported a different impact on CBP and indices, which were reported as significantly lower in the amlodipine-based regimen, despite the identical reduction of SBP and PP produced by the two regimens. Furthermore, this differencewas sustained over the 4-year follow-up.
The effect of ACE inhibition on pulsatile hemodynamics in patients with stable coronary disease and preserved left ventricular function was assessed in the PEACE (Prevention of Events with Angiotensin Converting Enzyme inhibition) substudy, which evaluated the response to trandolapril administration, regarding its effect on CBP and arterial stiffness. The carotid-femoral PWV was moderately decreased, but beyond expectations according to BP reduction or differences in baseline characteristics alone. Of note, no modification in aortic compliance, Aix, or total arterial compliance was reported.
The outcomes of the vascular substudy of the ANBP2 (second Australian National Blood Pressure Study) in an elderly hypertensive population had been much anticipated, due to the fact that, for the first time, the recording of CBP was performed before randomization in a large clinical trial. A subset (479 patients) of the trial cohort was evaluated both before and after 4 years of treatment. Although ANBP2 reported better prognosis for male (no beneficial effect was evident in women) elderly (65–84 years old) hypertensive subjects randomly assigned to the ACEI enalapril compared with a diuretic-based regimen, no differential effect on CBP and PBP was observed between the two arms. These results may be partly explained by the increased arterial stiffness, which occurs with age and hypertension.
In addition to the aforementioned long-term important studies, several others, most of them mid- (less than 6 months' duration) or short- (less than 4 weeks' duration, including acute effect studies lasting few hours) term, evaluated the effect of ACEIs (either as monotherapy, or as a single add-on treatment, or even head-to-head comparison). These studies provided convincing evidence that the reduction observed in arterial stiffness (as measured by PP, PWV, and AIx) and the positive effect on BP amplification should be considered a class effect (Table I). Furthermore, evidence clearly supports the presence of an additional CBP-lowering ability by ACEIs, beyond PBP, although this effect is not present in all the available studies. The authors proposed that the differential effect ofACEIs onCBP may be explained on the basis of their binding affinity for tissue ACE (e.g. quinapril has higher affinity than captopril) and their effect on endothelial function. Despite their minor differences, ACEIs seem to act by decreasing wave reflections potentially attributable to chronic inverse remodelling of the small arteries leading to reduced reflection coefficients.
5.2 Angiotensin Receptor Blockers
The aforementioned data regarding the effects of ACE inhibition on pulsatile hemodynamics suggest that ARBs, affecting the same pressor system, might produce similar results (Table II). In patients with resistant hypertension (receiving at least three antihypertensive drugs at maximum dosage, including an ACEI), the addition of valsartan for 2 weeks significantly reduced AIx. Arandomized, crossover study in hypertensive patients compared valsartan with captopril. The two drugs equally reducedPWVand AIx. In addition, their combination resulted in further reductions in arterial stiffness. Most importantly, this effect remained significant even after BP adjustment, implying that different drugs that independently improve arterial wall properties may lead to a greater reduction in arterial stiffness when combined.
In a 4-week crossover study, losartan was compared with hydrochlorothiazide (HCTZ) in patients with essential hypertension. The authors showed that only losartan reduced PWV and aortic AIx and augmented PP amplification, for comparable BP reduction. Similar results were reported for valsartan. Telmisartan was evaluated in hypertensive patients with type 2 diabetes mellitus, and was shown to substantially reduce PWV and AIx compared with placebo.
In hypertensive patients, eprosartan significantly decreased both PBP and CBP compared with baseline, as evidenced by the PP amplification ratio (peripheral/central PP ratio: 1.38 vs 1.41, p < 0.005), which was consistent with a CBP reduction beyond the PBP reduction. Of note, a decrease in absolute PP amplification was reported after eprosartan administration (15mmHg vs 13mmHg at baseline), creating reasonable doubt regarding the efficacy of eprosartan in reducing CBP beyond PBP. Nevertheless, wave reflections (AIx) and aortic stiffness (PWV) were significantly reduced. Olmesartan produced similar results in a different clinical trial when administered to patients with mild to moderate essential hypertension.
In another study, 15 hypertensive patients were given irbesartan or lisinopril for 12 weeks and then crossed over for 12 weeks. Arterial stiffness was assessed by measuring PWV in the carotid-femoral (CF), carotid-radial (CR), and femoral dorsalis-pedis (FD). While irbesartan and lisinopril reduced PWV(CF) in the elastic arterial system, only irbesartan reduced PWV (CR) and PWV (FD), enhancing arterial compliance in both elastic and muscular arteries. The difference between treatments remained significant after SBP adjustment.
Whereas the direct association between arterial stiffness reduction and decreased CV risk has been efficiently documented with ACEIs, it remains unclear with ARBs. However, important equivalence studies support the potential hypothesis that ARBs may reduce primary CV outcomes due to beneficial effects that are beyond PBP lowering.
In the LIFE (Losartan Intervention For Endpoint reduction in hypertension) study, losartan was shown to be superior to atenolol in preventing CV outcomes, but the effect was more pronounced than expected from the decrease in PBP. It has been speculated that this protective effect might be attributable to direct additional properties of losartan on arterial wall structure. Furthermore, in a small randomized, doubleblind, crossover study with a similar population to LIFE, it was shown that BNP was significantly lower in the losartan arm compared with atenolol after a 4-month treatment, indicating that the survival benefit reported for losartan in the LIFE study could be attributed to the observed improvement of BNP levels. Nevertheless, CBP and indices were not recorded and the authors reported no significant difference between losartan and atenolol in terms of AIx and PWV.
The VALUE (Valsartan Antihypertensive Long-Term Use Evaluation) study, which compared the potential effects of a valsartan-based regimen to an amlodipine-based regimen with regard to their effects on cardiac morbidity and mortality in hypertensive individuals at high risk of cardiac events, reported no difference in cardiac endpoints between the two treatment groups. However, in a subanalysis of the VALUE study, in which patients had to receive monotherapy with either of the two study drugs, despite similar BP reductions in the two subgroups, relative risks of heart failure and new-onset diabetes were in favor of the valsartan-based treatment. Finally, in the OPTIMAAL (Optimal Trial In Myocardial Infarction with the Angiotensin Antagonist Losartan) study, losartan was shown to share with captopril equally beneficial effects in reducing all-cause mortality in high-risk patients after myocardial infarction. Of note, pulsatile hemodynamics (i.e. CBP and indices, AIx and PWV) were not assessed in the aforementioned equivalence studies; consequently, no direct linking may be presumed between their results and possible modifications in arterial wall properties.
Pleiotropic effects have been attributed to both drug classes. Thus, it seems relevant to propose that ARBs may have beneficial effects on CBP beyond PBP lowering. The available data on their potential long-term effect on arterial stiffness is still weak and further clinical evidence is required.
5.3 Aldosterone Antagonists
In a randomized, 1-month study in a hypertensive population comparing spironolactone to bendrofluazide, only spironolactone decreased PWV and AIx and these effects remained significant after adjustment for BP change. Another randomized, double-blind trial evaluated the administration of either spironolactone or HCTZ as monotherapy in older hypertensive patients for 6 months, with a significant reduction in BP, PP, and PWV, which was comparable in both groups. Spironolactone was also shown to improve arterial stiffness in early-stage chronic kidney disease.
In a direct head-to-head comparison, eplerenone was proven to share with amlodipine equally beneficial effects on SBP, PP, and PWV in an elderly population suffering from isolated systolic hypertension. For comparable BP reduction, eplerenone was associated with decreased vessel stiffness and collagen/elastin ratio and reduced circulating inflammatory markers, contrary to atenolol administration that was associated with an increase in resistant artery stiffness.
5.4 Renin Inhibitors
The hemodynamic effects of the first direct renin inhibitor (DRI), aliskiren, were only recently investigated. Arterial stiffness was assessed in adults with uncomplicated type 1 diabetes mellitus, via recording AIx and PWV before and after aliskiren administration, during clamped euglycemia and hyperglycemia. The carotid AIx during clamped euglycemia was shown to decrease in parallel to PWV during clamped hyperglycemia, suggesting that DRIs improve arterial wall properties. Nevertheless, further clinical research is required.
5.5 Calcium Channel Blockers
The CAFE study, which was discussed in section 5.1, provided the most important evidence on the effect of dihydropyridine CCBs (amlodipine) on pulsatile hemodynamics. Despite the lack of baseline data regarding CBP and central indices, CAFE is singled out as the only available large study relating CBP modifications due to antihypertensive treatment to mortality.
A prospective study evaluated the effects of a CCB (azelnidipine) versus a diuretic (HCTZ), both in combination with the same ARB (olmesartan). Patients who were administered olmesartan for 3 months were randomized to receive either azelnidipine or HCTZ for another 6 months. The authors found a significantly greater reduction in central SBP in the olmesartan/azelnidipine group, despite the fact that brachial SBP was equally reduced in both groups. The decrease in AIx was similar between the two groups; however, when adjusted for a heart rate (HR) of 75 beats/min, the reduction was significantly greater in the olmesartan/azelnidipine group. PP amplification in the olmesartan/azelnidipine group was significantly increased as well. Furthermore, a more significant decrease in aortic stiffness, expressed by the aortic PWV, was reported in the olmesartanazelnidipine arm. These results suggest a superiority of azelnidipine over HCTZ, when combined with the same dose of olmesartan, regarding their effect on central hemodynamics, although no difference between peripheral SBP reductions was reported.
Several studies have evaluated the effect of dihydropyridine CCBs on arterial stiffness (Table III). Nitrendipine was shown to significantly reduce both PBP and CBP in patients with endstage renal disease. However, the substantially enhanced PP amplification suggests a more pronounced effect on central PP. Nitrendipine was also associated with reduced PWV and AIx.
CBP and AIx were reduced when patients with newly diagnosed, mild to moderate hypertension were administered barnidipine for a 6-month period. However, no substantial changes in intrinsic arterial stiffness were observed.
In another study in untreated hypertensive patients, felodipine and amlodipine were shown to reduce CBP to a similar extent. Even though both peripheral and central SBP were reduced, the authors reported a more pronounced effect on central rather than peripheral SBP, compared with placebo, as well as on PP. Of note, the aforementioned effects were shown to be dose-dependent, since they were enhanced with a 10mg dosage (compared with 5 mg) of either drug. The wave reflections were also significantly reduced (−10%).
Relatively few studies have directly compared the four major antihypertensive drug classes (ACEIs, CCBs, BBs, and diuretics) on PBP, CBP, and arterial stiffness. In one randomized study, 59 elderly patients suffering from untreated isolated systolic hypertension were assigned to one of the following antihypertensive drugs: perindopril, atenolol, lercanidipine, or bendrofluazide. Central and peripheral BP, pressure wave reflections and aortic stiffness were assessed at baseline, following 2 weeks of placebo therapy, and at the end of 10 weeks of active therapy. Peripheral SBP and peripheral PP were reduced to a similar extent by all four drug classes. Furthermore, all but atenolol reduced central PP. Lercanidipine was the only agent that significantly reduced AIx, whereas atenolol increased it, though this effect was not significant after adjustment for HR. Aortic PWV remained unchanged by all four drugs.
Apart from dihydropyridine CCBs, verapamil was shown to reduce CBP, PP, and PWV in hypertensive patients.
Several of the aforementioned studies compared the effects of CCBs with other antihypertensive drugs on CBP and PP amplification, suggesting that CCBs are superior to BBs and diuretics, but inferior to ACEIs.
5.6 Diuretics
Despite the fact that diuretics are reported to increase arterial compliance, up to date data suggest that this effect is mainly attributed to the decline in BP rather than to a change in the intrinsic properties of the arterial wall.
The vascular substudy of the ANBP2 trial provided noteworthy evidence regarding CBP modifications in response to administration of diuretics. Sixty-four percent of the participants were treated with HCTZ (the final choice belonged to the family practitioner). At the conclusion of the 4-year followup period, both PBP and CBP were significantly decreased in the HCTZ arm; however, the effect was more pronounced peripherally than centrally (SBP −17/−15mmHg, PP −9/−6mmHg, respectively).
The effect of standard (25 mg) and high-dose (50 mg) HCTZ on pulsatile hemodynamics was evaluated in elderly untreated patients with essential hypertension. Both doses reduced PBP and CBP as well as PP to a similar extent compared with placebo. Consequently, SBP and PP amplification remained unchanged at the end of the study, suggesting that HCTZ affects central and peripheral pressure parameters equally.
Bendrofluazide significantly reduced PBP compared with placebo, but had no impact on CBP. Comparison of spironolactone with bendrofluazide showed that only spironolactone reduced PWV and AIx and these effects remained significant even after adjustment for PBP change. Both indapamide and bendrofluazide altered hemodynamic parameters, reducing MAP and AIx to a similar extent in patients with stroke. Indapamide was also evaluated in subjects with essential hypertension for 2 months and no difference was observed regarding the effect on PBP, CBP, and AIx.
The above-mentioned studies reveal a neutral rather than a favorable effect of diuretics on pulsatile hemodynamics (Table IV). Consequently, diuretics are now considered to be inferior to the newer antihypertensive agents (ACEIs, CCBs) that have an additional effect on CBP beyond PBP, enhancing BP amplification. To date there are no available data from drug comparison trials involving diuretics and ARBs. In contrast, diuretics, due to their neutral effect on CBP, seem to outclass certain BBs (i.e. atenolol and bisoprolol), which were found to reduce BP amplification.
5.7 β-Blockers
Most of the available studies regarding the effect of BBs on pulsatile hemodynamics have used the cardioselective agent atenolol. The REASON study was the first clinical trial to provide clear proof that atenolol, despite reducing both CBP and PBP, had a more pronounced peripheral than central effect. Interestingly, central PP was actually slightly increased, although peripheral PP was substantially reduced. Of note, PWV was decreased by almost 1 msec, whereas AIx was increased by approximately 2.5%, which might be related to the HR deceleration by 8 beats/min. The aforementioned findings were consistent with a number of smaller short- to midterm studies, confirming the hypothesis that atenolol negatively affects both SBP and PP amplification and wave reflections, resulting in an increase in AIx. Instead, aortic stiffness, assessed by PWV, was significantly reduced.
The CAFE study, in agreement with the REASON study, provided sufficient evidence of the inferiority of atenolol to amlodipine, also relating the negative effects of atenolol on CBP and indices to adverse CV outcomes. Furthermore, a more pronounced effect was observed on PBP than CBP. Based on these results, meta-analyses have challenged the recommendation for classic BBs (namely atenolol and bisoprolol) as an optimal treatment for uncomplicated hypertension.
Whether such observations should be considered a class effect remains unclear. BBs represent a heterogeneous antihypertensive drug class (Table V). There is now accumulating evidence that newer BBs with peripheral vasodilator properties, such as nebivolol and carvedilol, may be more effective in improving central pulsatile measurements. In animalmodels, nebivolol, but not atenolol, was shown to improve arterial distensibility. In another study, nebivolol, in contrast to atenolol, increased PP amplification and reduced AIx, possibly due to increased NO levels that affect the small resistance arteries. On the contrary, another study showed that although atenolol increased central PP and AIx more than nebivolol, both drugs reduced PP amplification. When nebivolol was compared with enalapril in hypertensive diabetic patients, both agents were shown to possess beneficial properties regarding arterial compliance and glucose metabolism. In patients suffering from coronary artery disease, endothelial function was improved by nebivolol, though not by atenolol, following therapy for a month. In a recent, retrospective study, three groups of hypertensive patients treated with ARBs, carvedilol/nebivolol, or atenolol were compared. For similar PBP and PWV reduction, treatment with either vasodilator BBs or ARBs was associated with lower CBP and AIx compared with atenolol. Furthermore, atenolol significantly increased peripheral AIx compared with metoprolol, for comparable MAP reduction. Metoprolol showed non-inferiority to valsartan with regard to their effect on PWV in mildly hypertensive patients.
Further large-scale, long-term outcome studies are required to determine whether BBs, classic and/or newer, are associated with adverse clinical outcomes. Finally, although precautions exist against using classic BBs in hypertensive patients with multiple metabolic risk factors, such recommendations do not apply to the newer vasodilator BBs, at least to date.
5.8 Nitrates
Although nitrates were never considered as first-line antihypertensive treatment for essential hypertension, invasive studies assessing their acute effect on pulsatile hemodynamics have provided noteworthy data. In this context, two invasive studies reported a greater reduction in CBP compared with PBP after 5 minutes of low-dose nitrates, nitroglycerine (sublingually), or nitroprusside (intravenously). An increase in SBP amplification was observed in both studies. By contrast, AIx was significantly decreased.
In patients with isolated systolic hypertension, SBP and PP amplification were increased and AIx was substantially reduced after the administration of isosorbide mononitrate. Equally beneficial results were also reported in a small (n = 6) group of patients under long-term therapy with either captopril or eprosartan, suggesting that an alternative mechanism, other than RAAS inhibition, might be involved, potentially through the direct relaxation of the muscular conduit arteries. On the contrary, their effect on PWV was apparently inadequate or, at least, equivocal, implying that large arterial stiffness is minimally affected by nitrates
Source...