Endothelial Vascular Function in Hypertensive Patients After Renin—Angiotensin System Blockad

Abstract

It is unclear whether single and combined pharmacologic inhibition of the renin‐angiotensin‐aldosterone system have similar effects on endothelial function and blood pressure (BP). The authors evaluated 63 hypertensive patients divided into 4 groups (hydrochlorothiazide 25 mg/d; irbesartan [IRBE] 150 mg/d; quinapril [QUIN] 20 mg/d; or IRBE 150 mg/d + QUIN 20 mg/d) and 25 healthy normotensive subjects (normal) followed for 12 weeks. Endothelium‐dependent dysfunction measured as flow‐mediated dilation at Weeks 0 and 12 were: normal, 11.5%±2.4% vs 13.5%±2.0%; hydrochlorothiazide, 7.3%±2.0% vs 12.8%±3.1%; QUIN, 7.2%±2.8% vs 13.2%±2.1%; IRBE, 7.1%±2.8% vs 13.0%±2.9%; and IRBE + QUIN, 7.5%±1.9% vs 12.8%±3.0%. Nitroglycerin‐mediated responses were: normal, 26.0%±1.9% vs 24.0%±2.5%; hydrochlorothiazide, 17.0%±2.2% vs 18.3%±2.6%; QUIN, 17.8%±3.2% vs 23.4%±3.0%; IRBE, 16.8%±3.6% vs 24.7%±2.0%; and IRBE + QUIN, 17.3%±3.0% vs 25.1%±2.5%. Antihypertensive therapy restored BP to normal and improved the endothelium‐dependent and ‐independent dysfunction after renin‐angiotensin‐aldosterone system blockade. In a further finding, the combined effect of angiotensin‐converting enzyme inhibition and angiotensin II type 1 receptor blockade was not superior to the action of either of these treatments separately.

Hyperreactivity of the renin‐angiotensin‐aldosterone system (RAAS) is a risk factor for patients with primary hypertension. ¹ , ² It is well known that the normal endothelial surface has an intrinsic ability to prevent cardiovascular disease, ³ and vascular dysfunction precedes morphologic and structural changes associated with systemic arterial hypertension, including tissue remodeling and myocardial and vascular hypertrophy. ⁴ , ⁵ Abnormal endothelium‐dependent and smooth muscle relaxation has been reported in patients with essential hypertension and is associated with altered nitric oxide (NO) production and signal transduction ⁶ , ⁷ , ⁸ linked to an abnormal balance of the RAAS.

Several studies have shown that inhibition of the RAAS has a beneficial effect beyond the reduction in blood pressure. ⁹ , ¹⁰ These benefits include altering the hemodynamic action of angiotensin II as well as nonhemodynamic effects. ¹¹ , ¹² Angiotensin‐converting enzyme (ACE) inhibitors and angiotensin receptor blockers (ARBs) counteract the untoward effects of angiotensin II by targeting the RAAS. ⁹ Both ACE inhibition and blockade of angiotensin II type 1 (AT₁) receptors can prevent or reverse the action of angiotensin II in endothelial dysfunction. ¹³ , ¹⁴ It is uncertain, however, whether a simultaneous RAAS double blockade has benefits over a single blockade.

Hydrochlorothiazide (HCTZ) is the thiazide diuretic most commonly used in the treatment of hypertension. ¹⁵ Thiazide‐type diuretics have been the basis of antihypertensive therapy in most trials. ¹⁶ , ¹⁷ Thiazide‐type diuretics have produced remarkable results in lowering blood pressure and reducing morbidity and mortality associated with hypertension. ¹⁶ , ¹⁸ Diuretics initially reduce extracellular volume and cardiac output. Over time, these actions on volume become less important, although blood pressure remains lowered. Later actions of diuretics result largely from a reduction in peripheral vascular resistance; ¹⁹ however, it is unknown whether thiazide‐type diuretics alter endothelial function.

Recently it has become possible to assess vascular function by flow‐mediated dilation (FMD) using high‐frequency ultrasound imaging of the brachial artery. ²⁰ By this technique, we compared the effects on endothelial function in hypertensive patients of antihypertensive therapy with HCTZ and single or double blockade of the RAAS with an ACE inhibitor and/or an ARB.

Method

Study Population

This study had a randomized, open, and parallel 5‐group design in which conduit artery endothelium‐dependent and nitroglycerin (NTG)‐induced vasodilatation were assessed by FMD of the brachial artery (Figure). Two investigators who were not aware of the patients' treatment groups measured the vascular responses.

Figure.

Experimental protocol. FMD indicates flow‐mediated dilation; ABPM, ambulatory blood pressure monitoring.

The volunteers in this study were 25 normotensive subjects and 63 hypertensive patients with a diagnosis of primary hypertension in which the arterial blood pressure measured by ambulatory blood pressure monitoring (ABPM) was consistently >130/80 mm Hg (British Hypertension Society criterion). ²¹ , ²² The hypertensive participants were subdivided into 4 groups: diuretic (HCTZ 25–50 mg/d; n=18; 13 black/5 nonblack); ACE inhibitor (quinapril [QUIN] 20 mg/d; n=16; 9 black/7 nonblack); ARB (irbesartan [IRBE] 150 mg/d; n=14; 8 black/6 nonblack) and ACE inhibitor plus ARB (IRBE 150 mg/d + QUIN 20 mg/d; n=15; 7 black/8 nonblack). The 25 healthy subjects (NORM) were taking neither drugs nor placebo (n=25; 12 black/13 nonblack). Ethnicity was based on the physician's observation and the subject's self‐identification. Both normotensive and hypertensive individuals were examined every 2 weeks.

All of the hypertensive subjects were recruited from patients diagnosed in the outpatient clinic at the university hospital of the State University of Campinas (HC‐UNICAMP) and were followed up by specialists from the hypertension clinic at UNICAMP. All of the therapeutic regimens used were in agreement with current clinical standards and norms. All of the patients underwent common and specific laboratory tests, including abdominal ultrasound, renal scintigraphy, magnetic resonance imaging, and renal arteriography when necessary.

The exclusion criteria included secondary forms of hypertension, pheochromocytoma, renal artery stenosis, primary hyperaldosteronism, aortic coarctation, impaired renal function, ischemic heart disease, liver diseases, and other major diseases. Patients were also excluded if they had recently used medicines that affected vascular function, including statins, prostaglandin inhibitors, vitamins, contraceptives (within the previous 2 months), and acetylsalicylic acid (within the previous 7 days).

All groups of subjects provided a complete medical history and underwent a physical examination, an electrocardiogram, and laboratory analyses to exclude individuals with dyslipidemia (defined as low‐density lipoprotein cholesterol ≥159 mg/dL, according to the National Cholesterol Education Program Adult Treatment Panel III guidelines), ²³ diabetes mellitus (defined as overnight fasting glycemia >126 mg/dL, according to the American Diabetes Association, [ADA]), ²⁴ and evidence of hepatic, renal, or hematologic dysfunction.

This study was approved by the institutional Ethics Committee for Research on Humans and was carried out in accordance with current Brazilian legislation for human research. All of the patients were aware of the investigative nature of the study and gave informed written consent before their participation in the investigation.

Experimental Procedure

All procedures were performed in a quiet, air‐conditioned room (22°–24°C/72°–75°F), and were initiated at 8 AM after overnight fasting. At this time, baseline blood samples were drawn from all subjects from a forearm vein.

FMD Technique.

Vascular function was studied using ATL HDI ultrasound equipment (Advanced Technology Laboratories, Seattle, WA) and a high‐resolution 7–12 MHz linear vascular transducer coupled to computer‐assisted analysis software (Medware Sistemas Medicos Ltda, Brasília, DF, Brazil) and a computerized edge‐detection system for multiple measurements of the brachial artery diameter.

Arterial endothelial and vascular smooth muscle functions were assessed before and after the administration of antihypertensive therapies. Initially, the brachial artery responses to endothelial‐dependent (FMD) and ‐independent (NTG‐mediated) stimuli were determined using a modification of the technique introduced and described by Celermajer et al, ²⁵ and according to the guidelines for the ultrasound assessment of endothelium‐dependent FMD of the brachial artery as outlined in the report of the International Brachial Artery Reactivity Task Force published in 2002 by Correti et al. ²⁶ Twelve weeks later, the brachial artery FMD was again determined, after treatment with HCTZ, 25–50 mg/d, QUIN, 20 mg/d, IRBE, 150 mg/d, or a combination of IRBE 150 mg/d + QUIN, 20 mg/d. The follow‐up of 12 weeks was based on previous reports in which this period was considered adequate for the complete blockade of both ACE H3 and angiotensin receptors. ²⁷

NTG Technique.

The response to NTG was used as a measure of endothelium‐independent vasodilatation. ²⁸ After recording the second baseline scan, 0.4 mg of NTG Technique was given sublingually and 4–5 minutes later the brachial artery was imaged. The changes in brachial artery diameter in response to shear stress and NTG were expressed as the percentage change relative to the vessel diameter immediately before drug administration.

Systemic arterial blood pressure was recorded in all subjects using ABPM (SpaceLabs, Issaquah, WA), which automatically and intermittently measures blood pressure during a 24‐hour period. Blood pressure was measured before (Week 0) and after 12 weeks of oral medication.

Statistical Analyses

Descriptive data are expressed as means ± SDs. The sample size for shear stress‐mediated vasodilatation was calculated as a power of 0.70 (α=.s05; estimated difference = 10%; variability = 2%). This calculation was based on the guidelines published in 2002 by Correti et al. ²⁶

Variance analysis was used for blood pressure endothelium‐dependent and ‐independent responses (α=.05) to determine significant differences among the 5 groups (baseline and 12 weeks) as well as within each group (repeated measurements). Multiple comparison tests were used when necessary. Demographic data on the 5 groups were also compared by variance analysis.

Results

The baseline clinical characteristics of the hypertensive subjects are shown in Table I. There were no significant differences in relation to age, body mass index, glycemia, total cholesterol, low‐density lipoprotein cholesterol, triglycerides, sodium, potassium, plasma creatinine levels, and 24‐hour urinary sodium excretion.

Table I.

Baseline Characteristics of the Groups

	Control	hctz 25 mg/d	Quinapril 20 mg/d	Irbesartan 150 mg/d	Irbesartan 150 mg/s+ Quinapril 20 mg/d
Number	25	18	16	14	15
Age, y	46.6±9.0	49.4±7.9	48.8±8.6	50.3±7.5	49.9±5.1
Sex, men/women, No.	11/14	7/11	7/9	6/8	6/9
Body mass index, kg/m2	26.7±5.8	26.4±3.3	26.3±2.5	24.5±3.5	26.6±2.8
Systolic BP by ABPM, mm Hg	122.6±8.6	148±11	150±14	168±15	164±17
Diastolic BP by ABPM, mm Hg	79.9±4.9	92±9	94±11	90±12	90±11
Heart rate, bpm	73.3±11.4	68.0±7.4	72.2±5.8	71.9±6.4	72.9±5.2
Glycemia, mg/dL	84.7±11.2	90.9±6.3	92.6±10.0	91.5±5.7	91.9±5.1
Total cholesterol, mg/dL	191.0±26.6	212.7±11.4	206.0±35.7	200.4±39.4	232.0±31.9
LDL cholesterol, mg/dL	113±27.1	130.5±20.9	117.0±36.0	115.1±28.1	140.8±32.9
Triglycerides, mg/dL	114.1±54.9	151.1±42.9	141.1±46.1	174.5±81.9	200.6±52.8
Sodium, mEq/L	141.7±1.8	140.9±1.5	139.9±2.2	140.0±1.9	139.9±1.4
Potassium, mEq/L	4.2±0.5	4.1±0.4	4.0±0.3	4.1±0.2	4.0±0.2
Urine sodium, mEq/24 h	187.7±40.1	146.8±30.8	150.6±31.1	152.5±24.0	159.3±15.3
Creatinine, mg/dL	0.9±0.1	0.9±0.2	0.87±0.2	0.9±0.1	1.0±0.1
Data are presented as mean ± SD except where indicated. HCTZ indicates hydrochlorothiazide; ABPM, ambulatory blood pressure (BP) monitoring; and LDL, low‐density lipoprotein.

After 12 weeks of therapy, there was a significant reduction (P<.05) in the systolic and diastolic blood pressures in the 4 groups treated with antihypertensive agents when compared with Week 0 (Table II). There were no significant differences (P>.05) in relation to reduction of blood pressure among the hypertensive groups, regardless of whether mono‐therapy or combination therapy was employed.

Table II.

Results of 24‐Hour Ambulatory Blood Pressure (BP) Monitoring Before (Week 0) and After (Week 12) Antihypertensive Treatment

	Systolic BP, mm Hg		Dlastolic BP, mm Hg
Group	Week 0	Week 12	Week 0	Week 12
Normotensive	116±11	119±13	64±8	76±10
Hydrochlorothiazide	148±11*	123±15†	92±9*	79±12†
Quinapril	150±14*	117±16†	94±11*	76±10†
Irbesartan	168±15*	136±8†	90±12*	71±10†
Irbesartan + quinapril	164±17*	129±20†	90±11*	80±13†
Data are presented as mean ± SD. *P<.0001 vs normotensive. †P<.0001 vs Week 0.

The antihypertensive treatment with HCTZ, ACE inhibitor QUIN, angiotensin II type 1 receptor blocker IRBE, and ACE combined with ARB (IRBE + QUIN) promoted improvement in the endothelium‐dependent FMD (induced by shear stress and endogenous releasing of NO) after 12 weeks of treatment (Table III).

Table III.

Relative Changes in Diameter of the Brachial Artery in Responses to Flow‐Mediated Dilation and Nitroglycerin at 0 and 12 Weeks

	Brachial Artery Diameter, % Change
	Flow‐Mediated Dilation		Nitroglycerin
Group	Week 0	Week 12	Week 0	Week 12
Normotensive	11.5±2.4	13.5±2.0	26.0±1.9	24.0±2.5
Hydrochlorothiazide	7.3±2.0*	12.8±3.1†	17.0±2.2*	18.3±2.6‡
Quinapril	7.2±2.8*	13.2±2.1†	17.8±3.2*	23.4±3.0§
Irbesartan	7.1±2.8*	13.0±2.9†	16.8±3.6*	24.7±2.0§
Irbesartan ± quinapril	7.5±1.9*	12.8±3.0†	17.3±3.0*	25.1±2.5§
Data are presented as mean ± SD percentage changes. *P<.001 vs normotensive. †P<.001 vs flow‐mediated dilation at Week 0. ‡P<.05 vs quinapril, irbesartan, and irbesartan ± quinapril groups. §P<.001 vs nitroglycerin at Week 0.

The NTG‐mediated responses (shear‐stress independent, induced by NO donors) at Weeks 0 and 12 in the hypertensive groups were: NORM, 26.0%±1.9% vs 24.0%±2.5%; HCTZ, 17.0%±2.2% vs 18.3%±2.6%; QUIN, 17.8%±3.2% vs 23.4%±3.0%; IRBE, 16.8%±3.6% vs 24.7%±2.0% and IRBE + QUIN, 17.3%±3.0% vs 25.1%±2.5%. The NTG response at 12 weeks was significantly less with HCTZ than in the other groups (P<.05) (Table III).

Discussion

This prospective study was designed to compare the effects of a thiazide diuretic (HCTZ), an ACE inhibitor (QUIN) and an ARB (IRBE), alone or combined, on arterial endothelial dysfunction in patients with hypertension. In agreement with other studies, ⁶ , ⁷ , ²⁹ , ³⁰ FMD was reduced at baseline in the hypertensive patients studied here compared with normotensive individuals, but improved after 12 weeks of antihypertensive treatment.

As shown by Anderson, ³¹ administration of the ACE inhibitor QUIN improved FMD. Similarly, the ARB candesartan improved tonic NO release and reduced vasoconstriction to endogenous endothelin‐1 in the forearm of hypertensive patients, ³² whereas the ARB telmisartan did not improve endothelial dysfunction in hypertensive patients. With the exception of these studies, there are no other data on the effect of ARBs on the FMD of conduit arteries in patients with essential hypertension. ARBs can reverse endothelial dysfunction in patients with coronary artery disease with an effect similar to that of ACE inhibitors. ³³ Recently Klingbeil et al ³⁴ reported that AT₁ receptor blockade with valsartan improved basal NO production and release via a mechanism independent of blood pressure, concluding that long‐term treatment with IRBE enhanced endothelium‐dependent and ‐independent vasodilatation. In addition to this nonspecific effect, IRBE restored the vasoconstrictor capacity of NO synthase inhibitors, which suggested a direct effect on tonic NO release and decreased endothelin production. These actions may play an important role in the vascular protective effects of IRBE.

We found no cumulative antihypertensive effect when an ACE inhibitor and ARB were combined (IRBE + QUIN group) compared with treatment with each drug alone (IRBE or QUIN groups). In addition, 12 weeks of therapy with HCTZ was as efficacious in normalizing blood pressure as ACE inhibition and ATt blockade. ¹³ , ³⁵ , ³⁶ HCTZ diuretics restore only endothelium‐dependent dysfunction, however, with no effects on endothelium‐independent vasodilatation. This is a very interesting finding suggesting the beneficial action of HCTZ on vascular reactivity, but not on cardiac remodeling, in contrast to that seen with the other regimens.

Beneficial effects of RAAS blockade on blood pressure include the hemodynamic action of angiotensin II and nonhemodynamic effects. ¹¹ , ¹² Although ARBs produce beneficial vascular effects comparable to those of ACE inhibitors, ³⁷ , ³⁸ therapeutic doses of ACE inhibitors can produce incomplete inhibition of ACE activity and still allow sufficient conversion of angiotensin I to maintain the plasma basal levels of angiotensin II. ¹¹ Thus, theoretically ACE inhibitors and ARBs counteract the untoward effects of angiotensin II by targeting the RAAS, ⁹ but this association can cause a broader range of nonspecific effects, including actions on ACE or an ACE‐independent pathway such as the chymostatin‐sensitive angiotensin II‐generating enzyme or chymase. ¹³ , ³⁹ , ⁴⁰ The results obtained in this study may be due to a reduction in the activity or a dysregulation of these latter mechanisms. The findings reported here strengthen the relevance of blood pressure reduction in treating hypertension because they indicate involvement of the RAAS and NO biosynthesis in endothelial dysfunction in this disease. For example, rheologic features (such as reduced shear stress and pulse pressure decreases) are also associated with blood pressure reduction. This hemodynamic equilibrium is in large part related to RAAS and endothelial interaction.

ARBs produce beneficial vascular effects in a different manner than ACE inhibitors. ³⁷ , ³⁸ The ACE inhibitors reduce angiotensin II levels in the plasma and extrarenal tissues, whereas plasma angiotensin II levels increase in response to AT₁ receptor blockade alone. In contrast, renal angiotensin II levels do not increase in response to the AT₁ receptor blockers. The inhibition of angiotensin II formation in processes mediated by the kidneys may therefore have a central role in mediating the effects of ACE inhibitors and ARBs on blood pressure and in other structural and functional effects in the cardiovascular system. ⁴⁰ , ⁴¹ Combined ACE inhibition and the blockade of AT₁ receptors prevent the increase in plasma angiotensin II levels seen with ARBs alone, ¹² , ⁴² , ⁴³ , ⁴⁴ , ⁴⁵ and may be more effective in modulating angiotensin II‐mediated physiologic responses than treatment with either drug alone. ⁴⁶ Thus, the association of captopril with valsartan has a combined effect in reducing the arterial wall thickness of hypertensive patients. ⁴⁷ In rats, the neointimal spreading characteristically seen after damage of the arterial vessel wall with a catheter balloon is significantly attenuated by combined treatment with an ACE inhibitor (benazepril) and ARB (valsartan), probably because of an increase in NO bioavailability and a decrease in vascular oxidative stress. ⁴⁸ Recent clinical data indicate that combinations of ACE inhibitors and ARBs have a combined effect in reducing blood pressure and proteinuria. ⁴⁹ Plasma aldosterone levels show a variable, transitory increase during therapy with ACE inhibitors and ARBs alone or in combination (termed aldosterone escape phenomenon) ¹ , ² , ⁵⁰ via a mechanism that is NO‐dependent. ⁵¹ Aldosterone has a central role in cardiovascular remodeling following endothelial dysfunction in hypertensive patients treated with ACE inhibitors and ARBs. ¹ , ²

Our results indicate that endothelial dysfunction may be reversed by antihypertensive treatment with IRBE and QUIN, either alone or in combination. Moreover, the antihypertensive effect of these treatments was superimposed. Future studies using the same regimen of treatment can clarify the relationship between the functional results we found and morphologic features of cardiovascular remodeling in hypertension.

Conclusions

The results of this study showed that treatment with HCTZ, IRBE, QUIN, and IRBE + QUIN for 12 weeks normalized blood pressure and restored vascular reactivity (endothelium‐dependent and ‐independent vasodilatation) in hypertensive patients treated with ACE inhibitor, ARB, and combination thereof (single and dual blockade of the RAAS), but HCTZ restored only endothelial‐dependent dysfunction and did not improve endothelial‐independent vasodilatation. In addition, a combination of ACE inhibition and ATX blockade was not superior to either of these blockades alone.

Disclosure:

This work was supported by Fundacdo de Amparo a Pesquisa do Estado de Sao Paulo. Dr Moreno is supported by a research fellowship from Conselho National de Desenvolvimento Cientifico e Tecnologico.