Journal of Human Hypertension (2007) 21, 654–663 & 2007 Nature Publishing Group All rights reserved 0950-9240/07 $30.00 www.nature.com/jhh ORIGINAL ARTICLE Acute effects of renin-angiotensin system blockade on arterial function in hypertensive patients
KA Aznaouridis, KS Stamatelopoulos, EN Karatzis, AD Protogerou, CM Papamichael and JP Lekakis Vascular Laboratory, Department of Clinical Therapeutics, Athens Medical School, Alexandra Hospital, Athens, Greece
The acute effects of the renin-angiotensin system (RAS) significant effect (P ¼ NS). Additionally, AIx was reduced blockers may be important in some clinical settings. To after quinapril (absolute decrease of 7.2%, Po0.01) and assess the acute impact of such drugs on arterial marginally after captopril (decrease of 4.7%, P ¼ 0.07). function, we studied the effects of captopril 25 mg, Only quinapril led to a beneficial change of FMD quinapril 20 mg and telmisartan 80 mg on 100 hyperten- (absolute increase of 2.7%, Po0.001). No treatment sive patients, according to a randomized, double-blind, was related to significant changes of peak hyperaemic placebo-controlled study. Central (aortic) blood pres- or 3-min hyperaemic FBF. In adjusted analyses, all the sure (BP) and augmentation index (AIx, a measure of favourable alterations induced by quinapril were inde- wave reflections), as well as flow-mediated dilatation pendent of potential confounding haemodynamic fac- (FMD) of the brachial artery and forearm blood flow tors. Our data show that acute RAS inhibition with (FBF) (measures of conduit and resistance artery quinapril (20 mg) may be more beneficial in terms of endothelial function, respectively), were evaluated be- arterial function and central haemodynamics compared fore and 2 h after oral drug administration. Compared to to captopril (25 mg) or telmisartan (80 mg). Further placebo, captopril and quinapril decreased central studies are needed to investigate whether these acute systolic (by 7.5 mm Hg, Po0.05 and by 12.3 mm Hg, arterial effects of quinapril are clinically significant. Po0.001) and diastolic BP (by 4.9 mm Hg, Po0.01 and Journal of Human Hypertension (2007) 21, 654–663; by 8.4 mm Hg, Po0.001), whereas telmisartan had no doi:10.1038/sj.jhh.1002211; published online 26 April 2007
Keywords: endothelium; wave reflection; ACE inhibition; central blood pressure; antihypertensive drugs
Introduction fully explained by the respective BP change.8–10,12,14 However, different ACEIs may not confer a same Arterial elastic properties, central (aortic) haemody- degree of organoprotection, and this is perhaps due namics and peripheral endothelial function are to dissimilarities in their propensity to penetrate important predictors of cardiovascular risk.1–4 Im- vascular tissue and inhibit the tissue ACE.15,16 portantly, the recent Conduit Artery Function Although the notion of a class effect of ACEIs has Evaluation (CAFE´ ) study showed that a greater been proposed, this is not fully documented or decrease of central blood pressure (BP) with anti- universally accepted. On the other hand, angioten- hypertensive treatment is associated with reduced sin-II type-1 receptor blockers (ARBs) efficiently cardiovascular events, independent of peripheral BP block the interaction of both ACE- and non-ACE- changes.5 Likewise, there are data showing that the produced angiotensin-II with type-1 receptor, and reversal of endothelial dysfunction, that is present 6 7 these drugs are also characterized by organoprotec- in hypertensive patients, may benefit prognosis. tive effects.17 Drugs that block the renin-angiotensin system Beyond the established benefit of long-term anti- (RAS), such as angiotensin converting enzyme hypertensive therapy, there are no data to support (ACE) inhibitors (ACEIs), may improve cardiovas- that acute effects of drugs – including RAS blockers cular structure and function,8–13 and this effect is not – on BP and arterial function are clinically relevant in hypertensive patients. However, the acute effects Correspondence: Dr K Aznaouridis, Department of Cardiology, of RAS blockers may be important in some other Athens Medical School, Kyparissias 14, Kato Acharnes, Athens clinical settings, such as acute myocardial infarc- 13671, Greece. tion,18–20 coronary revascularization19,21 and even E-mail: [email protected] 22 Received 3 January 2007; revised 20 March 2007; accepted 21 non-cardiac surgery. Furthermore, there is evi- March 2007; published online 26 April 2007 dence that different drugs of the same class (ACEIs) Acute RAS blockade and arterial function KA Aznaouridis et al 655 may not be equally beneficial in such acute condi- trolled room at 231C. After a 20-min rest period, tions.21,23,24 Data comparing the acute vascular baseline measurements for evaluation of central effects of different RAS blockers in humans are haemodynamics (pulse-wave analysis), of endothe- limited.16,25–27 Accordingly, in the present rando- lial function of resistance and then of conduit mized, placebo-controlled, double-blind, parallel- arteries were taken in the supine position, in this design study, we investigated the acute effects of fixed order. The study of conduit vessels was different types of RAS blockade on arterial function preceded by a 15-min rest period to allow recovery in a population with impaired arterial performance, of vascular function. Then, the subjects were using a thorough approach that evaluates endothe- randomized to take either captopril 25 mg or lial function of both conductance and resistance quinapril 20 mg or telmisartan 80 mg, or placebo arteries and central (aortic) haemodynamics. For per os, together with drinking 200 ml of water. this purpose, we compared the effects of captopril, Randomization was undertaken by sealed envelopes quinapril (an ACEI with presumed high-tissue and gave rise to four groups with 25 patients each. affinity15,16) and telmisartan (an ARB possessing a Dosing was supervised. In a pilot study that unique action profile28) on subjects with essential consisted of eight patients in each treatment arm hypertension, which may be regarded as a model who underwent BP measurement every 30 min after of abnormal arterial function. Also, we sought to drug administration, we observed that the three investigate whether any observed vascular changes active drugs caused a maximal decrease of BP would be associated with respective alterations of approximately at 2 h. Therefore, vascular studies peripheral BP. were repeated 2 h after drug intake in all groups. The study protocol was approved by our Institu- tional Research Ethics Committee and all subjects Materials and methods gave informed consent. Study population We evaluated consecutive patients with mild to Evaluation of central haemodynamics (pulse-wave moderate hypertension who were referred from the analysis) Hypertension Unit to our Laboratory for vascular Central (aortic) BPs and augmentation index (AIx), studies for research purposes. A full medical history an index of wave reflections,1,3,29–31 were calculated was taken and physical examination was performed. using a validated, commercially available system Patients who had evidence of secondary hyperten- (SphygmoCor, AtCor Medical, Sydney, Australia), sion, coronary artery disease, heart failure, history of which employs the principle of applanation tono- a cerebrovascular event, endocrinopathy, or an acute metry. Waveforms of radial pressure (provided by or chronic inflammatory-infectious disease were radial artery tonometry) were calibrated according to excluded. All subjects were clinically well and sphygmomanometric SBP and DBP measured in the taking no antioxidant vitamins, anti-inflammatory brachial artery, since there is practically negligible or steroid substances. No female participant was on pressure amplification between the brachial and oral contraceptives or oestrogen replacement ther- radial arteries. Mean arterial pressure (MAP) was apy. The final study comprised 100 patients (mean then computed automatically by numerically aver- age 57.2 years, 48 males). aging of the radial pressure waveform.32 The central Brachial systolic and diastolic BP (SBP and DBP) BP was derived with the use of a generalized transfer in the sitting position were measured in the right function, which is an accurate estimate of the arm with a mercury sphygmomanometer on three central arterial pressure waveform. Augmented occasions 1 min apart, after the subjects had rested pressure (AP) is the pressure added to the incident for 15 min, and the mean value was calculated. wave by the returning reflected one and represents Hypertension was diagnosed when BP was above the pressure boost with which the left ventricle must 140/90 mm Hg in three different visits a week apart, cope at systole. AIx (defined as AP divided by pulse or if chronic use of antihypertensive drugs was pressure and expressed as a percentage) is a documented. In that case, medication was with- composite measure of the magnitude of wave drawn for 2 weeks before the study. Subjects reflections and arterial stiffness, which affects abstained from caffeine, ethanol and flavonoid- timing of wave reflections. For similar heart rate containing beverages for at least 12 h before the and effective length of the arterial system, larger study. values of AIx indicate increased wave reflections from the periphery and/or earlier return of the reflected wave as a result of increased pulse-wave Study design velocity (owing to increased arterial stiffness), and The study was carried out using a randomized, vice versa.1,3,31 The pressure at the inflection point double-blind, placebo-controlled, parallel-group de- of the aortic waveform (P1) corresponds to the peak sign. Subjects were studied in the morning after an of blood flow velocity and is a measure of the force overnight fast on two occasions, before and 2 h after of cardiac ejection. Arrival time (Dt) of reflected drug administration, in a quiet, temperature-con- waves at aorta is the time from the foot of the
Journal of Human Hypertension Acute RAS blockade and arterial function KA Aznaouridis et al 656 pressure wave to the late systolic peak and repre- Ten minutes after the 2-h cuff deflation, endothe- sents the time needed for pressure waves to travel lium-independent, nitrate-induced dilatation (NID) from the aorta to peripheral arterial sites and return was measured after delivering a single (0.4 mg) dose back to the aorta owing to wave reflections. A lower of nitroglycerin spray sublingually. Images were Dt indicates a shorter travel time of the pressure recorded on super-VHS (Video Home System) waves and a higher arterial stiffness.30 Absolute videotape and were measured offline by the same amplification of pulse pressure between central observer, who was blinded to the image sequence and peripheral arteries (in mm Hg) was calculated and the randomization assignment. Three cardiac as brachial pulse pressureÀaortic pulse pressure, cycles were analysed and measurements were and relative amplification (in %) was calculated as averaged. The mean variability for FMD measured ((brachial pulse pressureÀaortic pulse pressure)/ on two different days by this observer was 1.1% brachial pulse pressure)  100. (absolute value). Brachial artery flow was calculated according to the equation: Flow (in ml/min) ¼ mean flow veloci- Evaluation of endothelial function in resistance vessels ty  heart rate  3.14  (brachial artery diameter/2)2. The response of forearm blood flow (FBF, expressed Reactive hyperaemia was expressed as the percent as ml/min per 100 ml of forearm tissue volume) to change of brachial artery flow from baseline: reactive hyperaemia is a non-invasive method of hyperaemia (in %) ¼ ((hyperaemic flowÀresting estimating the endothelium-dependent vasodilation flow)/resting flow)  100. Shear stress (averaged of the small resistance arteries.14 FBF studies were over the whole cardiac cycle) was calculated using performed in the right arm using a strain gauge the equation: Shear stress (in dyn/cm2) ¼ venous occlusive plethysmograph (EC-5R, Hokan- 8  m  mean flow velocity/resting diameter, where son DE Inc., WA, USA). Mercury-filled silastic strain m is the viscosity of blood.35 FMD was calculated as gauges of appropriate size for each subject were the percent change of brachial artery diameter from placed about 5 cm below the antecubital crease, at baseline: FMD (%) ¼ ((hyperaemic diameterÀresting the level of the maximal circumference of the diameter)/resting diameter)  100. forearm. A wrist cuff was inflated to 50 mm Hg above the SBP throughout the study to exclude the hand circulation. An upper arm rapid cuff inflator Statistical analysis (E20, Hokanson DE Inc., WA, USA), which was Sample size calculations were based on data from inflated to 50 mm Hg for 7 s in each a 15 s cycle, was our unit, which showed that the s.d. of AIx and FMD used to occlude intermittently venous outflow from for hypertensive subjects were 8 and 2.4%, respec- the arm. After resting FBF was obtained, a cuff tively. Therefore, we estimated that 22 subjects per placed on the mid-forearm was inflated to supra- group would provide 80% power at the 5% level of systolic levels (50 mm Hg above SBP), and the significance to detect a difference of 7% in AIx in a arterial inflow was thus interrupted. After 4.5 min, parallel design study. Similarly, 24 subjects per this cuff was deflated leading to reactive hyperae- group would provide 80% power to detect an mia. FBF was then measured for 3 min. Resting absolute difference of 2% in FMD. forearm vascular resistance (FVR) was calculated as Continuous variables are expressed as mean the MAP divided by resting FBF. value7s.d., whereas categorical variables as abso- lute and/or relative frequencies. All continuous variables were tested for homogeneity of variance Evaluation of endothelial function in conductance and normal distribution before any statistical analysis vessels was applied, by using the Kolmogorov–Smirnov Resting and hyperaemic diameters and flows as well criterion. For skewed variables, data are expressed as flow-mediated dilatation (FMD) of the conduit as median value (25th–75th percentile) and loga- brachial artery, an estimate of endothelial function, rithmic transformation was performed before analy- were determined by using a linear array ultrasonic sis. Within each group, comparisons of values after transducer (Acuson 128 XP, Mountain View, CA, treatment with baseline values were carried out with USA) as described previously.33,34 Two-dimensional paired t-test. The unpaired t-test or one-way analysis and Doppler images of the right brachial artery for of variance (ANOVA) was used for comparisons diameter and flow determination were initially among groups. In cases that ANOVA yielded a recorded under resting conditions. Resting mean significant difference among treatments, isolation flow velocity was assessed by a pulsed Doppler of the active treatment(s) that produced the differ- signal, from a sample volume placed in the middle ence compared to placebo was carried out by the part of the artery. Then, reactive hyperaemia was Dunnett post hoc test. The changes of vascular induced as described above. Brachial artery was parameters were adjusted for baseline values or continuously scanned from 30 s before to 90 s after other potential confounders with analysis of covar- cuff deflation. Hyperaemic mean flow velocity was iance (ANCOVA). Contingency tables and w2 test or measured within the first 15 s and hyperaemic the Fisher’s exact test were applied for categorical diameter was measured 50–60 s after cuff release. parameters. Comparisons of the time course curves
Journal of Human Hypertension Acute RAS blockade and arterial function KA Aznaouridis et al 657 of FBF during reactive hyperaemia before and after (brachial) BP differentiated significantly across the treatment within each group were carried out by four treatment groups (Figure 1). Compared with two-way repeated-measures ANOVA (2 treatment placebo, captopril and quinapril significantly de- points (baseline and after treatment) Â 13 time creased both SBP (by 10.2 mm Hg, Po0.01 and by points (one FBF measurement each 15 s for 3 min)). 10.4 mm Hg, Po0.01) and DBP (by 3.9 mm Hg, Correlations between variables were evaluated by Po0.05 and by 8.6 mm Hg, Po0.001). On the other calculation of the Pearson correlation coefficient. hand, telmisartan decreased DBP (by 3.9 mm Hg, Exact Po0.05 were considered statistically signifi- Po0.05) but not SBP (decrease of 4.2 mm Hg, cant. Data analysis was performed with SPSS soft- P ¼ NS). Both captopril and quinapril decreased ware, version 10.1 (SPSS Inc., Chicago, IL, USA). MAP (by 6.5 mm Hg, Po0.01 and by 10.6 mm Hg, Po0.001, respectively), whereas the reduction of MAP achieved with telmisartan was of borderline Results significance (by 3.9 mm Hg, P ¼ 0.1). The effect of each active drug vs placebo is better Concerning central (aortic) pressures, captopril described by reporting response, defined as mean and quinapril decreased SBP (by 7.5 mm Hg, Po0.05 net active drug effect minus placebo effect. and by 12.3 mm Hg, Po0.001) and DBP (by 4.9 mm Hg, Po0.01 and by 8.4 mm Hg, Po0.001) compared with placebo (Figure 1), whereas telmisartan had no Baseline characteristics significant effect (decrease of 4.9 mm Hg in SBP and There were no significant differences in clinical of 3.1 mm Hg in DBP, both P ¼ NS). Interestingly, characteristics or baseline parameters of vascular with quinapril, the reduction of central and periph- function among the four study subgroups (Tables 1 eral SBP were similar (P ¼ NS); on the other hand, and 2). with captopril, the absolute decrease of aortic SBP Overall, women had higher AIx (33.777.9 vs was lower than the decrease of brachial SBP 25.4710.9%, Po0.001) and lower Dt (103.1710.5 (Po0.05). ANCOVA showed that the reduction of vs 108.0711.8 msec, Po0.05), absolute PP amplifi- central SBP and DBP were independent of the cation (8.775.6 vs 11.275.7 mm Hg, Po0.05) respective changes of peripheral BPs with quinapril and % PP amplification (16.579.2 vs 21.6710.6%, (both Po0.05), but not with captopril (both P ¼ NS). Po0.05) compared with men. Patients with previous The absolute mean change of AIx was similar antihypertensive treatment were older (59.079.5 vs in men and women (À2.7 vs À3.2%, P ¼ NS) and 54.378.2 years, Po0.05) and they had a marginally in patients with and without previous treatment lower peripheral DBP (90.2712.4 vs 95.0711.5 mm (À2.2 vs À4.1%, P ¼ NS) and correlated with the Hg, P ¼ 0.06) than never-treated patients. change of brachial SBP (r ¼ 0.27, Po0.01) and DBP (r ¼ 0.22, Po0.05). The four treatments differed significantly (Po0.01) with regard to the changes Heart rate, BP and pulse-wave analysis of AIx (Figure 2). Compared to placebo, AIx Heart rate did not change significantly with any decreased significantly with quinapril (absolute treatment (P ¼ NS). The changes of peripheral change by 7.2%, Po0.01), decreased marginally
Table 1 Clinical characteristics of the study population according to treatment
Captopril Quinapril Telmisartan Placebo (n ¼ 25) (n ¼ 25) (n ¼ 25) (n ¼ 25)
Age (years) 57.479.6 56.8710.4 56.378.9 58.478.5 Sex (male/female) 10/15 13/12 12/13 13/12 Smoking 7 (28%) 6 (24%) 7 (28%) 3 (12%) Dyslipidemia 4 (16%) 3 (12%) 7 (28%) 10 (40%) Diabetes 3 (12%) 1 (4%) 2 (8%) 2 (8%) Duration of EH (years) 2 (1–6.5) 5 (1–12) 2 (1–7) 4 (1–9) BMI (kg/m2) 28.074.0 28.174.2 28.373.9 30.674.9 Family history for CAD 9 (36%) 7 (28%) 13 (52%) 6 (24%) Previous treatment 16 (64%) 15 (60%) 14 (56%) 17 (68%) ACEI 4 (16%) 5 (20%) 3 (12%) 8 (32%) ARB 4 (16%) 4 (16%) 2 (8%) 6 (24%) Calcium antagonist 9 (36%) 6 (24%) 6 (24%) 6 (24%) b-blocker 4 (16%) 6 (24%) 7 (28%) 7 (28%) Diuretic 5 (20%) 6 (24%) 5 (20%) 9 (36%)
Abbreviations: ACEI, angiotensin converting enzyme inhibitor; ANOVA, analysis of variance; ARB, angiotensin-II type-1 receptor blocker; BMI, body mass index; CAD, coronary artery disease; EH, essential hypertension. Categorical variables are presented as absolute and/or relative frequencies, while continuous variables as mean7s.d. or median (25th–75th percentile). For comparisons among the four groups, all P-values 40.05 (derived from w2 test or Fisher’s exact test or one-way ANOVA).
Journal of Human Hypertension Acute RAS blockade and arterial function KA Aznaouridis et al 658 Table 2 Baseline vascular parameters of the study population according to treatment
Captopril Quinapril Telmisartan Placebo (n ¼ 25) (n ¼ 25) (n ¼ 25) (n ¼ 25)
HR (beats/min) 74.1716.8 67.9710.4 68.1710.9 72.0713.0 Peripheral SBP (mm Hg) 148.8718.9 146.6715.9 148.7719.2 147.8719.3 Peripheral DBP (mm Hg) 89.479.4 92.9712.2 95.9715.3 89.8710.8 Peripheral PP (mm Hg) 59.3716.5 53.7710.4 52.8713.3 58.0717.6
Pulse-wave analysis Central SBP (mm Hg) 136.8715.6 136.1714.6 139.0716.3 133.1716.7 Central DBP (mm Hg) 92.479.3 94.2713.2 95.8714.5 89.779.5 Central PP (mm Hg) 44.5713.4 41.979.5 43.2712.1 43.4716.0 MAP (mm Hg) 111.7710.6 111.8713.7 114.4714.0 108.5710.6 AP (mm Hg) 13.076.3 12.676.1 14.476.6 13.278.3 AIx (%) 29.0711.0 29.279.8 32.4710.4 28.2710.1 7 7 7 7 P1 (mm Hg) 31.6 10.9 29.1 5.3 29.0 7.7 30.7 8.8 Dt (msec) 104.8715.9 106.8710.0 105.578.5 104.7710.3 PP amplification (mm Hg) 11.475.7 9.874.4 9.674.9 8.877.6 PP amplification (%) 20.478.5 19.278.2 18.578.4 17.6714.6
Brachial artery study Resting diameter (mm) 4.1270.60 4.4170.72 4.2870.70 4.4370.65 Hyperaemic diameter (mm) 4.2870.66 4.5370.71 4.3870.68 4.5470.63 FMD (%) 3.7773.41 2.7571.87 2.4972.14 2.6572.30 Resting flow (ml/min) 103 (64–168) 69 (50–132) 81 (57–163) 92 (60–138) Hyperaemic flow (ml/min) 225 (139–299) 187 (139–334) 186 (135–287) 220 (141–338) Hyperaemia (%) 1467157 1827137 1617123 144796 Resting shear stress (dyn/cm2) 7.874.5 6.774.4 6.973.3 6.373.8 Hyperaemic shear stress (dyn/cm2) 15.777.7 14.577.1 15.176.5 13.075.9 Hyperaemic change of shear stress (%) 1467142 1677133 1537122 130792
Forearm blood flow study Resting FBF (ml/min/100 ml of tissue) 3.671.1 3.971.3 4.571.4 4.071.4 Resting FVR (mm Hg/ml/min/100 ml of tissue) 34.0712.3 32.3714.5 29.1716.4 30.3710.6 Peak hyperaemic FBF (ml/min/100 ml of tissue) 7.474.3 8.473.8 9.373.0 7.773.9 Peak hyperaemia (%) 113794 1357123 124799 102796
Abbreviations: AIx, augmentation index; ANOVA, analysis of variance; AP, augmented pressure; DBP, diastolic blood pressure; Dt, arrival time of reflected wave; FBF, forearm blood flow; FMD, flow-mediated dilation; FVR, forearm vascular resistance; HR, heart rate; MAP, mean arterial pressure; PP, pulse pressure; SBP, systolic blood pressure. Variables are presented as mean7s.d. or median (25th–75th percentile). For comparisons among the four groups, all P-values 40.05 (derived from one-way ANOVA).
Figure 1 Box-and-whisker plots of changes at 2 h (differences from baseline) in peripheral (a) and central (b) systolic and diastolic blood pressure, according to treatment. The centerline of the box denotes the median value of the change, the extremes of the box, the interquartile range and the bars, the upper and lower limits of 95% of the data. The circles represent outlying data (1.5–3.0 times the interquartile range beyond the 25th or 75th percentile). P-values by one-way ANOVA. wPo0.05, zPo0.01, yPo0.001 (Dunnett post hoc test comparing changes between each active drug group with placebo group). ANOVA, analysis of variance.
with captopril (by 4.7%, P ¼ 0.07) and did not the change of MAP in a subsequent ANCOVA change with telmisartan (absolute decrease of model, quinapril was still associated with a sig- 2.0%, P ¼ NS). After introducing baseline AIx and nificant reduction of AIx compared to placebo
Journal of Human Hypertension Acute RAS blockade and arterial function KA Aznaouridis et al 659
Figure 2 Box-and-whisker plots of aortic augmentation index (AIx, left plot) and flow-mediated dilation of the brachial artery (FMD, right plot) at baseline and after 2 h, according to treatment. P-values within groups by student’s t-test for paired measures. P-values among groups (top of graphs) by one-way ANOVA comparing differences from baseline. ANOVA, analysis of variance.
(Po0.05), whereas, on the other hand, the marginal able effect of quinapril was due to a significant association between captopril and change of AIx increase of brachial hyperaemic diameter (Po0.05) disappeared (P ¼ NS). with quinapril (responses compared to placebo of Compared to placebo, we found a greater but 0.012, 0.144 and 0.028 mm after captopril, quinapril nonsignificant increase of Dt with captopril (by and telmisartan, respectively). No significant differ- 3.3 msec) and quinapril (by 2.8 msec) than with ences (P ¼ NS) were observed among treatments telmisartan (by 1.6 msec), indicating a trend for with regard to resting brachial diameter (responses reduction of arterial stiffness with ACEIs. We also of 0.004, 0.028 and 0.016 mm), resting flow, hyper- observed a greater increase of absolute and relative aemic flow and flow velocity, degree of reactive PP amplification with quinapril (by 2.1 mm Hg and hyperaemia or hyperaemic shear stress (responses of 6.4%) than with captopril (change of À0.1 mm Hg À1.9, À2.1 and 0.3 dyn/cm2). NID values were and 2.6%) or telmisartan (change of À0.5 mm Hg similar among the four groups at the end of the and 0.6%), suggesting a higher reduction of left study (9.774.3 vs 12.475.9 vs 10.475.1 vs ventricular afterload with quinapril, but this differ- 10.174.3% with placebo, captopril, quinapril and ence was not statistically significant among groups. telmisartan, respectively, P ¼ NS). P1 did not change with any treatment (P ¼ NS), indicating no effects on the force of cardiac ejection. FBF study Compared with placebo, treatment with the three Brachial artery study active drugs was not associated with significant The change of FMD after treatment was not different changes (P ¼ NS) of resting FBF (responses of À0.1, between men and women (absolute mean increases À0.2 and 0.1 ml/min per 100 ml of tissue after of 0.80 vs 1.36%, P ¼ NS), whereas we observed a captopril, quinapril and telmisartan, respectively). trend for a lower change in patients who were taking Likewise, resting FVR did not change significantly drugs before the study compared with never-treated with treatment (responses of 1.0, À2.2 and À2.2 mm patients (absolute mean increases of 0.78 vs 1.59%, Hg/ml/min per 100 ml of tissue, P ¼ NS). Further- P ¼ 0.1). Analysis revealed an inverse association of more, peak hyperaemic FBF remained unaltered the increase of FMD with baseline FMD (r ¼À0.40, (P ¼ NS) after treatment (responses of 0.1, À0.5 and Po0.001), but no correlation with age or the change 0.6 ml/min per 100 ml of tissue). Finally, we did not of BPs (P ¼ NS). observe any significant change of hyperaemic FBF There was a significant effect of treatment on FMD throughout the 3-min period of hyperaemia after among the four groups (Po0.001, Figure 2). Post hoc treatment with the three active drugs or placebo (all analysis showed that this effect was attributed to an P ¼ NS with two-way repeated-measures ANOVA, increase of FMD with quinapril (absolute increase of Figure 3). 2.70%, Po0.001 compared with placebo). The effects of captopril and telmisartan were not sig- nificant (absolute increases of 0.35 and 0.21%, Discussion respectively, both P ¼ NS). ANCOVA showed that the association between quinapril treatment and To our knowledge, this is the first study to compare FMD increase was independent of baseline FMD directly the acute vascular effects of different types value or previous treatment (Po0.001). This favour- of RAS blockade, by using a thorough approach that
Journal of Human Hypertension Acute RAS blockade and arterial function KA Aznaouridis et al 660
Figure 3 Three-minute net hyperaemic forearm blood flow (hyperaemic values minus resting value) before and 2 h after treatment. P-values by two-way repeated measures ANOVA. ANOVA, analysis of variance.
integrates measures of endothelial function of both effects of high tissue affinity ACEIs are more conduit and resistance arteries, and central (aortic) favourable compared to other types of RAS block- haemodynamics. Our data indicate that quinapril ade. We also confirm a previous study showing that and captopril may decrease central BP in an acute acute BP reduction with captopril does not improve setting. Moreover, quinapril also exerts favourable FMD.41 On the other hand, two studies have shown effects on central AIx and conduit artery endothelial that a single dose of captopril may improve function. Notably, these actions of quinapril are resistance artery vasodilation in hypertensives, but independent of accompanying peripheral BP in both studies intrabrachial agonists were used to changes; rather, they seem to stem from properties stimulate the forearm endothelium,42,43 so direct that are inherent to quinapril. comparisons with our study cannot be made.
Endothelial function Central haemodynamics Quinapril acutely increased the vasodilatory reserve In our study, quinapril reduced AIx independent of of conduit arteries but not of resistance arteries, MAP change and also decreased central BPs, but whereas captopril and telmisartan lacked any this was not the case for captopril. In general, aortic significant effect. The favourable effect of quinapril AIx is determined by heart rate, by the force of on FMD is largely attributed to an increase of cardiac ejection, by the effective length of the brachial hyperaemic diameter, given that other arterial system, by the amount of the wave that is parameters that may interfere with FMD, such as reflected back to the aorta (that mainly depends on resting brachial diameter, reactive hyperaemia, the tone of the resistance arteries) and finally, by the shear stress or the ability of arterial smooth muscle velocity of the reflected wave (arterial stiff- to dilate (expressed by NID),33,35 were not affected by ness).1,3,5,31 As there was no significant difference
any treatment. Quinapril has a higher binding in P1 (an expression of cardiac contractility) or heart affinity for tissue ACE than captopril15,16 and may rate between treatment arms, we conclude that the increase bradykinin and nitric oxide (NO) more higher decrement of central SBP with quinapril was potently.15,36 NO is the main mediator of FMD33 and predominantly due to lower pressure wave augmen- a determinant of arterial elastic properties.37 Thus, tation due to decreased amount and/or delayed the observed effects of quinapril vs captopril and timing of pulse-wave reflection. telmisartan are perhaps due to improved NO Consistent with the decreased wave reflection, bioavailability. The NO pathway contributes more AIx was significantly lower with quinapril. to vasomotion of conduit arteries compared to Although our findings do not substantiate a vasodi- resistance arteries,38 so this is perhaps why quina- latory effect for quinapril, at least in the forearm pril had different effects on the brachial artery vessels (resting FBF and FVR, which are measures of compared with the forearm vasculature. the resting tone of small forearm arteries, did not Our results are in line with other acute, short-term change), we cannot certainly exclude a possible or chronic studies16,27,39,40 showing that the vascular vasodilatory effect on other vascular beds. In
Journal of Human Hypertension Acute RAS blockade and arterial function KA Aznaouridis et al 661 substance, this is most probably the case, since DBP guide chronic antihypertensive treatment, and they and MAP, which correlate with total peripheral do not suggest, by any means, that quinapril is a resistance, were reduced more with quinapril. This better antihypertensive drug. Rather, our observa- speculative differential effect is in line with evi- tions highlight the need for prospective randomized dence showing that oral quinapril inhibits acutely outcome trials in hypertensive populations to vascular ACE to a greater degree compared with investigate whether these differences in acute older ACEIs, despite a similar reduction of plasma arterial effects of RAS blockers are maintained in ACE.15,16 the long term, and more importantly, whether they Although we did not measure pulse-wave velocity translate into differences in clinical outcomes. in our study, we speculate that quinapril may have favourably affected the timing of wave reflection as well (due to a decrease in arterial stiffness), because (i) quinapril increased both Dt and PP amplification Specific comments – limitations (though nonsignificantly, since our sample size was Most patients were using antihypertensive drugs not powered to detect a significant change of these before the study. It is rather unlikely that our main parameters) and (ii) quinapril improved brachial results were influenced by previous treatment, since artery responses (FMD) mediated by NO, which is there were no differences in the frequency of an important determinant of arterial stiffness.37 This previous drug use across all four subgroups, and is in line with evidence showing that quinapril furthermore, we evaluated changes of variables from reduces predominantly the stiffness of muscular- baseline. Our subjects had abstained from their type arteries,44 such as the brachial artery and medications for 2 weeks, which is a longer interval abdominal aorta. than the recommended period of four drug half- Telmisartan blocks both ACE- and non-ACE- lives.33 However, we cannot certainly exclude a produced angiotensin II17, possesses some unique possible carry-over effect of previous drug use. features (such as an ability to activate the peroxi- The doses of drugs and the 2-h period were some proliferator–activated receptor-g28), and it has selected based on pharmacokinetics and the timing been shown to improve arterial elasticity, central of maximal BP drop in our pilot study. In other haemodynamics and endothelial function.45–47 How- studies that observed acute drug effects on arterial ever, in our acute study, we did not observe any function, these effects occurred almost simulta- significant effect of telmisartan on arterial function. neously with the BP change.16,50 However, because the vascular effects observed in our study were largely BP independent, we cannot preclude possi- Potential clinical implications ble delayed effects. Although data with regard to a potential clinical Dose may influence the chronic and even the significance of acute improvements of arterial func- acute vascular effects of some RAS blockers.8,50 We tion are lacking, there is some evidence that the used telmisartan 80 mg, since this dose may result in acute effects of RAS blockade may be important in a greater arterial effect compared with lower doses.50 some clinical settings, such as acute myocardial On the other hand, studies that compared the acute infarction,15,18–20 coronary revascularization19,21 and endothelial effects of different ACEIs showed that during major non-cardiac surgery.22 Furthermore, increasing the dose of the low-tissue affinity ACEI some data indicate that high tissue affinity ACEIs does not modify the effect on arterial function.25,26 may be more beneficial in terms of outcomes in such Furthermore, even a higher dose of captopril (50 mg) acute conditions18 partly because of more favourable does not improve FMD in hypertensive patients.41 effects on factors related to short term prog- Thus, the doses used in this study seem appropriate nosis.21,23,24,48 Our observations provide insights into to compare the three drugs. However, we cannot acute pharmacological actions of different RAS certainly rule out further improvements in acute blockers. However, the clinical relevance of the arterial effects with use of significantly higher doses. observed differences in acute arterial effects of RAS In conclusion, our study indicates an acute blockers, if any, is still speculative. favourable effect of oral quinapril on peripheral In hypertensive patients, the benefit of antihyper- and central arterial function. It suggests that in an tensive treatment is mainly determined by the acute setting, changes of peripheral BP do not reflect extent of BP reduction49 and the long-term effect reliably the effect of treatment on vascular perfor- on cardiovascular structure and function.8,9,11–14,39 mance. Moreover, it seems there are differences Recent data support that quinapril and captopril are between drugs that act on different points of the equally efficient in modifying important surrogate RAS, and even between drugs of the same class but end points with prognostic significance, such as left with different properties regarding their vascular ventricular mass.11 On the other hand, some studies effects. We need further studies to determine have underscored that an improvement of arterial whether the observed effects of the different RAS function in hypertensives may benefit prognosis,5,7 blockers are clinically relevant to chronic treatment but there are no trials to directly compare different of hypertensives or to conditions in which an acute ACEIs with regard to outcomes. Our findings cannot inhibition of the RAS is warranted.
Journal of Human Hypertension Acute RAS blockade and arterial function KA Aznaouridis et al 662 pressure or beyond the brachial artery? J Hypertens What is known on this topic 2005; 23: 551–556. K Changes of peripheral blood pressure do not accurately reflect 11 Aznaouridis K, Vyssoulis G, Karpanou E, Marinakis A, the effects of antihypertensive treatment on vascular performance. Barbetseas J, Zervoudaki A et al. Left ventricular (LV) K Blockers of the renin-angiotensin system may have beneficial geometry and dipping state are determinants of LV effects on arterial function and cardiovascular risk even in mass reduction with ACE-inhibitor antihypertensive acute settings. treatment. Blood Press Monit 2007; 12: 87–94. 12 Uehata A, Takase B, Nishioka T, Kitamura K, Akima T, What this study adds Kurita A et al. Effect of quinapril versus nitrendipine K The acute vascular effects of blockers of the renin-angiotensin on endothelial dysfunction in patients with systemic system are independent, at least in part, of the respective hypertension. Am J Cardiol 2001; 87: 1414–1416. acute changes of peripheral blood pressure. 13 Tsang TS, Barnes ME, Abhayaratna WP, Cha SS, Gersh K Administration of quinapril (20 mg) is associated with a more BJ, Langins AP et al. Effects of quinapril on left atrial favourable acute effect on conduit artery endothelial function and central haemodynamics compared with captopril (25 mg) structural remodeling and arterial stiffness. Am J or telmisartan (80 mg). Cardiol 2006; 97: 916–920. 14 Higashi Y, Sasaki S, Nakagawa K, Ueda T, Yoshimizu A, Kurisu S et al. A Comparison of angiotensin- converting enzyme inhibitors, calcium antagonists, References beta-blockers and diuretic agents on reactive hyper- emia in patients with essential hypertension: a multi- 1 O’Rourke MF, Adji A. An updated clinical primer on center study. J Am Coll Cardiol 2000; 35: 284–291. large artery mechanics: implications of pulse wave- 15 Dzau VJ, Bernstein K, Celermajer D, Cohen J, Dahlof B, form analysis and arterial tonometry. Curr Opin Deanfield J et al. Pathophysiologic and therapeutic Cardiol 2005; 20: 275–281. importance of tissue ACE: a consensus report. Cardi- 2 Laurent S, Boutouyrie P, Asmar R, Gautier I, Laloux B, ovasc Drugs Ther 2002; 16: 149–160. Guize L et al. Aortic stiffness is an independent 16 Lyons D, Webster J, Benjamin N. Effect of enalapril and predictor of all-cause and cardiovascular mortality quinapril on forearm vascular ACE in man. Eur J Clin in hypertensive patients. Hypertension 2001; 37: Pharmacol 1997; 51: 373–378. 1236–1241. 17 Schiffrin EL. Vascular and cardiac benefits of angioten- 3 Vlachopoulos C, Aznaouridis K, Stefanadis C. Clinical sin receptor blockers. Am J Med 2002; 113: 409–418. appraisal of arterial stiffness: the Argonauts in front of 18 Wienbergen H, Schiele R, Gitt AK, Juenger C, Heer T, the Golden Fleece. Heart 2006; 92: 1544–1550. Meisenzahl C, et al., for the MITRA PLUS Study 4 Widlansky ME, Gokce N, Keaney JF, Vita JA. The Group. Impact of ramipril versus other angiotensin- clinical implications of endothelial dysfunction. JAm converting enzyme inhibitors on outcome of unse- Coll Cardiol 2003; 42: 1149–1160. lected patients with ST-elevation acute myocardial 5 Williams B, Lacy PS, Thom SM, Cruickshank K, infarction. Am J Cardiol 2002; 90: 1045–1049. Stanton A, Collier D et al. CAFE Investigators; Anglo- 19 Lazar HL, Bao Y, Rivers S, Bernard SA. Pretreatment Scandinavian Cardiac Outcomes Trial Investigators; with angiotensin-converting enzyme inhibitors attenu- CAFE Steering Committee and Writing Committee. ates ischemia-reperfusion injury. Ann Thorac Surg Differential impact of blood pressure-lowering drugs 2002; 73: 1522–1527. on central aortic pressure and clinical outcomes: 20 Hikosaka M, Yuasa F, Yuyama R, Motohiro M, Mimura principal results of the Conduit Artery Function J, Kawamura A et al. Effect of angiotensin-converting Evaluation (CAFE) study. Circulation 2006; 113: enzyme inhibitor on cardiopulmonary baroreflex sen- 1213–1225. sitivity in patients with acute myocardial infarction. 6 Park JB, Charbonneau F, Schiffrin EL. Correlation of Am J Cardiol 2000; 86: 1241–1244. endothelial function in large and small arteries in 21 Lazar HL, Bao Y, Rivers S, Colton T, Bernard SA. High human essential hypertension. J Hypertens 2001; 19: tissue affinity angiotensin-converting enzyme inhibi- 415–420. tors improve endothelial function and reduce infarct 7 Modena MG, Bonetti L, Coppi F, Bursi F, Rossi R. size. Ann Thorac Surg 2001; 72: 548–554. Prognostic role of reversible endothelial dysfunction in 22 Kurzencwyg D, Filion KB, Pilote L, Nault P, Platt RW, hypertensive postmenopausal women. J Am Coll Rahme E et al. Cardiac medical therapy among patients Cardiol 2002; 40: 505–510. undergoing abdominal aortic aneurysm repair. Ann 8 Tropeano AI, Boutouyrie P, Pannier B, Joannides R, Vasc Surg 2006; 20: 569–576. Balkestein E, Katsahian S et al. Brachial pressure- 23 Tsikouris JP, Suarez JA, Simoni JS, Ziska M, Meyerrose independent reduction in carotid stiffness after long- GE. Exploring the effects of ACE inhibitor tissue term angiotensin-converting enzyme inhibition in penetration on vascular inflammation following diabetic hypertensives. Hypertension 2006; 48: 80–86. acute myocardial infarction. Coron Artery Dis 2004; 9 Asmar RG, London GM, O’Rourke MF, Safar ME, For 15: 211–217. the REASON Project coordinators and investigators. 24 Tsikouris JP, Suarez JA, Meyerrose GE, Ziska M, Fike Improvement in blood pressure, arterial stiffness and D, Smith J. Questioning a class effect: does ACE wave reflections with a very-low-dose perindopril/ inhibitor tissue penetration influence the degree of indapamide combination in hypertensive patient. A fibrinolytic balance alteration following an acute comparison with atenolol. Hypertension 2001; 38: myocardial infarction? J Clin Pharmacol 2004; 44: 922–926. 150–157. 10 Hirata K, Vlachopoulos C, Adji A, O’Rourke MF. 25 Houben AJ, Kroon AA, de Haan CH, Fuss-Lejeune MJ, Benefits from angiotensin-converting enzyme inhibitor de Leeuw PW. Quinaprilat-induced vasodilatation ‘beyond blood pressure lowering’: beyond blood in forearm vasculature of patients with essential
Journal of Human Hypertension Acute RAS blockade and arterial function KA Aznaouridis et al 663 hypertension: comparison with enalaprilat. Cardio- converting enzyme inhibition in vivo. Differential vasc Drugs Ther 2000; 14: 657–663. contribution of nitric oxide and bradykinin in con- 26 Hornig B, Arakawa N, Haussmann D, Drexler H. ductance and resistance arteries. Circulation 1996; 93: Differential effects of quinaprilat and enalaprilat on 1734–1739. endothelial function of conduit arteries in patients 39 Ghiadoni L, Magagna A, Versari D, Kardasz I, Huang Y, with chronic heart failure. Circulation 1998; 98: Taddei S et al. Different effect of antihypertensive 2842–2848. drugs on conduit artery endothelial function. Hyper- 27 Kimura M, Umemura K, Kosuge K, Nishimoto M, tension 2003; 41: 1281–1286. Ohashi K, Nakashima M. Attenuation by ACE inhibitor 40 Anderson TJ, Elstein E, Haber H, Charbonneau F. drugs of alpha-adrenoceptor sensitivity in human Comparative study of ACE-inhibition, angiotensin II vessels: possible differences related to drug lipophili- antagonism, and calcium channel blockade on city. Br J Clin Pharmacol 1998; 46: 599–603. flow-mediated vasodilation in patients with coronary 28 Sharma AM. Telmisartan The ACE of ARBs? Hyperten- disease (BANFF study). J Am Coll Cardiol 2000; 35: sion 2006; 47: 1–2. 60–66. 29 Vlachopoulos C, Dima I, Aznaouridis K, Vasiliadou C, 41 Ghiadoni L, Huang H, Magagna A, Buralli A, Taddei S, Ioakeimidis N, Aggeli C et al. Acute systemic inflam- Salvetti A. Effect of acute blood pressure reduction on mation increases arterial stiffness and decreases wave endothelial function in the brachial artery of patients reflections in healthy individuals. Circulation 2005; with essential hypertension. J Hypertens 2001; 19: 112: 2193–2200. 547–551. 30 Karatzis E, Papaioannou TG, Aznaouridis K, Karatzi K, 42 Hirooka Y, Imaizumi T, Masaki H, Ando S, Harada S, Stamatelopoulos K, Zampelas A et al. Acute effects of Momohara M et al. Captopril improves impaired caffeine on blood pressure and wave reflections in endothelium-dependent vasodilation in hypertensive healthy subjects: should we consider monitoring patients. Hypertension 1992; 20: 175–180. central blood pressure? Int J Cardiol 2005; 98: 425–430. 43 Millgard J, Hagg A, Sarabi M, Lind L. Captopril, but not 31 Segers P, De Backer J, Devos D, Rabben SI, Gillebert TC, nifedipine, improves endothelium-dependent vasodi- Van Bortel LM et al. Aortic reflection coefficients and lation in hypertensive patients. J Hum Hypertens 1998; their association with global indexes of wave reflection 12: 511–516. in healthy controls and patients with Marfan’s syn- 44 Topouchian J, Brisac AM, Pannier B, Vicaut E, Safar M, drome. Am J Physiol 2006; 290: H2385–H2392. Asmar R. Assessment of the acute arterial effects of 32 Safar ME, Blacher J, Pannier B, Guerin AP, Marchais converting enzyme inhibition in essential hyperten- SJ, Guyonvarc’h PM et al. Central pulse pressure and sion: a double-blind, comparative and crossover study. mortality in end-stage renal disease. Hypertension J Hum Hypertens 1998; 3: 181–187. 2002; 39: 735–738. 45 Asmar R, Gosse P, Topouchian J, N’tela G, Dudley A, 33 Corretti MC, Anderson TJ, Benjamin EJ, Celermajer D, Shepherd GL. Effects of telmisartan on arterial stiffness Charbonneau F, Creager MA et al. International in Type 2 diabetes patients with essential hyperten- Brachial Artery Reactivity Task Force. Guidelines for sion. J Renin Angiotensin Aldosterone Syst 2002; 3: the ultrasound assessment of endothelial-dependent 176–180. flow-mediated vasodilation of the brachial artery. JAm 46 Vingerhoedt NM, Gilles R, Howes JB, Griffin M, Howes Coll Cardiol 2002; 39: 257–265. LG. Haemodynamic and pulse wave responses to 34 Papamichael CM, Aznaouridis KA, Karatzis EN, intravenous infusions of angiotensin II during chronic Karatzi KN, Stamatelopoulos KS, Vamvakou G et al. telmisartan therapy in normal volunteers. J Renin Effect of coffee on endothelial function in healthy Angiotensin Aldosterone Syst 2003; 4: 244–248. subjects: the role of caffeine. Clin Sci (London) 2005; 47 Morimoto S, Yano Y, Maki K, Sawada K. Renal and 109: 55–60. vascular protective effects of telmisartan in patients 35 Mitchell GF, Parise H, Vita JA, Larson MG, Warner E, with essential hypertension. Hypertens Res 2006; 29: Keaney JF et al. Local shear stress and brachial artery 567–572. flow-mediated dilation. The Framingham Heart Study. 48 Wojewodzka-Zelezniakowicz M, Chabielska E, Mogiel- Hypertension 2004; 44: 134–139. nicki A, Kramkowski K, Karp A, Opadczuk A et al. 36 Napoli C, Sica V, de Nigris F, Pignalosa O, Condorelli Antithrombotic effect of tissue and plasma type M, Ignarro LJ et al. Sulfhydryl angiotensin-converting angiotensin converting enzyme inhibitors in experi- enzyme inhibition induces sustained reduction of mental thrombosis in rats. J Physiol Pharmacol 2006; systemic oxidative stress and improves the nitric oxide 57: 231–245. pathway in patients with essential hypertension. Am 49 Staessen JA, Wang JG, Thijs L. Cardiovascular preven- Heart J 2004; 148: e5. tion and blood pressure reduction: a quantitative 37 Wilkinson IB, MacCallum H, Cockcroft JR, Webb DJ. overview updated until 1 March 2003. J Hypertens Inhibition of basal nitric oxide synthesis increases 2003; 21: 1055–1076. aortic augmentation index and pulse wave velocity in 50 Stangier J, Su CA, van Heiningen PN, Meinicke T, van vivo. Br J Clin Pharmacol 2002; 53: 189–192. Lier JJ, de Bruin H et al. Inhibitory effect of telmisartan 38 Sudhir K, Chou TM, Hutchison SJ, Chatterjee K. on the blood pressure response to angiotensin II Coronary vasodilation induced by angiotensin- challenge. J Cardiovasc Pharmacol 2001; 38: 672–685.
Journal of Human Hypertension