Comparison of Effects of Angiotensin I– Converting Enzyme Inhibition and β-Blockade for 2 Years on Function of Small Arteries From Hypertensive Patients

The generation of increased peripheral resistance, which is the hallmark of high blood pressure in experimental animals and in humans, is the result of the altered structure and function of resistance arteries.¹²³ These are small arteries and arterioles of less than 400 μm and less than 100 μm in lumen diameter, respectively. Small arteries from hypertensive patients present numerous structural and functional changes⁴⁵ that are similar to those found in small arteries of hypertensive experimental animals.²³⁶⁷⁸⁹ The altered structure of small arteries is characterized by a thicker media in hypertensive patients than in normotensive subjects, with an increased media width–to–lumen diameter ratio. This increase in the ratio between the media thickness and the lumen diameter is the result of either vascular hypertrophy, in which the wider media encroaches on the lumen and reduces its diameter, or remodeling, in which the outer diameter of the vessel is reduced.¹⁰ The cellular bases of vascular hypertrophy and remodeling are controversial.¹¹¹²¹³ Whether by reduction of the outer diameter or by encroachment of the thicker media on the lumen, the increase in the media-to-lumen ratio resulting from either remodeling or hypertrophy will amplify vascular responses to most vasoconstrictors.¹⁴ This amplification may be one of the main phenomena underlying enhanced responsiveness of small arteries in hypertensive humans and experimental animals. Indeed, and in contrast to what is usually thought, media stress development in response to most vasoconstrictors such as norepinephrine, serotonin, vasopressin, and endothelin-1 is attenuated in small arteries from hypertensive patients.⁴⁵¹⁵ In addition, it has been repeatedly reported that hypertensive patients present varying degrees of impairment of endothelial function, namely endothelium-dependent relaxation elicited by acetylcholine,¹⁶ although some investigators have been unable to find evidence of depressed responses to acetylcholine in hypertensive patients.¹⁷

A study of the effects of antihypertensive treatment was performed with two specific agents—cilazapril, an angiotensin I–converting enzyme inhibitor marketed in Canada and Europe, and the β-blocker atenolol—to investigate the effects of these drugs on small arteries from patients with mild essential hypertension. The results of treatment with these drugs on the structure and function of small arteries after 1 year¹⁸ and on the structure of these vessels after 2 years¹⁹ have been previously reported. In the current study we compared the effect of 2 years of these antihypertensive therapies on vasoconstrictor responses and endothelium-dependent relaxation of small arteries from hypertensive patients.

Methods

Patients

The protocol was approved by the Ethics Committees of the Clinical Research Institute and Hôtel-Dieu Hospital of Montreal. All subjects gave written informed consent to participate in the study. Normotensive control subjects and patients with essential hypertension aged 25 to 50 years were recruited. All subjects studied were male because of requirements of the Health Protection Branch, Health and Welfare Canada, in relation to the potential side effects of drugs being studied if female participants became pregnant during the 2-year trial. Control subjects had a systolic blood pressure <140 mm Hg and diastolic blood pressure <85 mm Hg. Recumbent systolic blood pressure of hypertensive patients was >140 mm Hg and recumbent diastolic blood pressure was >90 mm Hg on more than two occasions. Most patients were newly diagnosed as having hypertension, and except for three patients were previously untreated. The three had not received antihypertensive medication for at least 6 months when they entered the study. The diagnosis of essential hypertension was established by the absence of clinical evidence of secondary hypertension; normal serum electrolytes, creatinine, urinalysis; a normal abdominal echogram; and, when indicated, renal scintiscan, renal arteriogram, and computed abdominal tomography. Patients were excluded if they smoked more than five cigarettes per day, if their fasting blood glucose level was abnormal, if their serum creatinine concentration was >150 μmol/L, or if they had any other systemic disease.

Subjects arrived at the hospital between 7:30 and 9:00 am, having fasted since the previous evening. Blood pressure (standard mercury sphygmomanometer) was measured after subjects had rested for 15 minutes in the sitting position. Diastolic blood pressure was read as phase V of the Korotkoff sounds. Twenty-four–hour ambulatory blood pressure monitoring was recorded at hourly intervals during daytime activities (8 am to 10 pm) in the hypertensive patients with a model 90207 Spacelabs ambulatory blood pressure recorder (Spacelabs Inc). Gluteal biopsies of subcutaneous fat measuring 1.0×0.5×0.5 cm³ were obtained, with the subjects under local anesthesia with 2% xylocaine, by the same surgeon throughout the study.

Treatment Protocol

Patients were randomly assigned to treatment with either atenolol or cilazapril in a double-blind fashion. They were seen at 2-week intervals twice before starting treatment, during which time they received placebo. Drug titration was done at 2-week intervals, with atenolol provided in identical 50- and 100-mg tablets and cilazapril in 2.5- and 5-mg tablets. If patients did not achieve the goal blood pressure (a diastolic blood pressure of 90 mm Hg or a reduction in diastolic blood pressure of 10 mm Hg), long-acting nifedipine was added at a dose of 10 or 20 mg twice daily. Only three patients required addition of nifedipine at the end of the titration period during the first year, and an additional patient started receiving nifedipine in the middle of the second year when his diastolic blood pressure remained consistently above 95 mm Hg. Patients were seen at monthly intervals during the rest of the first year. At the end of 1 year of treatment, the patients underwent a second biopsy of gluteal subcutaneous fat, as well as ambulatory blood pressure recording. During the second year, patients were seen at 3-month intervals. At the end of the second year, they underwent a third biopsy of gluteal subcutaneous fat, and ambulatory blood pressure recording was again performed.

Study of Small Subcutaneous Arteries

Arteries were dissected from the gluteal fat under a dissecting microscope immediately after the biopsy was obtained, and all small vessels found (up to 4 small arteries) were isolated.¹⁵¹⁸ Vessels were mounted as a ring preparation on an isometric myograph (Living Systems Instrumentation). The vessels were warmed to 37°C and allowed to equilibrate in physiological salt solution (PSS) (composition in mmol/L: NaCl 120, NaHCO₃ 25, KCl 4.7, KH₂PO₄ 1.18, MgSO₄ 1.18, CaCl₂ 2.5, EDTA 0.026, and glucose 5.5) for about 30 minutes with the vessel internal circumference set to give a wall tension of 0.2 mN/mm. The resting tension–internal circumference relationship was determined and vessels were set to L₀, where L₀=0.9 · L₁₀₀, L₁₀₀ being the internal circumference the vessels would have when relaxed and under a transmural pressure of 100 mm Hg. Measurements of vascular parameters were made at 12 sites along the vessel and averaged, as previously described.¹⁸ The vessels reported in this study had a calculated lumen diameter of 150 to 400 μm. They were maintained in PSS at 37°C for a further 90 minutes and were then stimulated as follows: (1) three stimulations (2 minutes each) with PSS, in which NaCl was substituted by KCl on an equimolar basis (K-PSS), and two stimulations with K-PSS containing 10 μmol/L norepinephrine; (2) two cumulative concentration-response curves to norepinephrine (from 0.01 to 10 μmol/L, 3 minutes per concentration); (3) a cumulative concentration-response curve to arginine vasopressin (from 0.01 to 30 nmol/L, 3 minutes per concentration); (4) a cumulative concentration-response to angiotensin II (from 0.1 to 300 nmol/L, 3 minutes per concentration); or (5) a cumulative concentration-response to endothelin-1 (from 0.01 to 1 μmol/L, 6 minutes per concentration). After each activation, the vessels were washed with PSS for 15 minutes. Relaxation of blood vessels precontracted with 10 μmol/L norepinephrine was performed with acetylcholine (1 nmol/L to 10 μmol/L).

Statistical Analysis

Results are presented as mean±SEM. pD₂, an index of the sensitivity of the complete concentration-response curves, was calculated as the negative log of EC₅₀. The EC₅₀ was the concentration (in mol/L) producing half the maximal response to an agent. Statistical comparisons were made by ANOVA for repeated measures or by one-way ANOVA followed by Duncan’s range statistic.

Results

The characteristics of the patient population and of normotensive subjects have been previously reported and are depicted in Table 1. The mean age of the eight patients randomly assigned to receive atenolol (age, 42.4±1.6 years) was similar to that of the nine who received cilazapril (age, 39.1±2.3 years), as was mean body mass index (Table 1). Blood pressure was well controlled throughout the 2 years of the study.

We previously reported that the response of small arteries of mildly hypertensive patients to endothelin-1 was blunted,¹⁵ and that after 1 year of cilazapril treatment, efficacy of the response improved, whereas under atenolol there was little change.¹⁸ Table 2 shows the maximum media stress developed in response to endothelin-1 by small arteries after 1 and 2 years of antihypertensive treatment as well as sensitivity of the complete concentration-response curves (represented by pD₂) to endothelin-1, and demonstrates a persistent correction in cilazapril-treated patients. Responses to norepinephrine, angiotensin II, and vasopressin were similar in all groups. Thus, after the second year of treatment, an improvement of vasoconstrictor responses to endothelin-1 similar to that seen after 1 year was found in the cilazapril-treated patients, confirming the reproducibility and persistence of this finding. As was found after 1 year of treatment, there was no change in the efficacy or any improvement in the sensitivity of the endothelin-1–induced constriction of small arteries in the atenolol group.

Endothelium-dependent relaxation of small arteries was investigated by examining the effect of acetylcholine on arteries precontracted with 10 μmol/L norepinephrine. Although, and as reported after 1 year of treatment,¹⁸ the concentration-response curves to acetylcholine were similar in both groups of patients and in normotensive subjects, small differences between groups could be detected when the degree of relaxation obtained at the highest concentrations of acetylcholine (1 and 10 μmol/L) was analyzed. The Figure shows that at the highest concentrations of acetylcholine used, relaxation of small arteries was slightly depressed in the hypertensive patients compared with the age-matched control subjects. As was observed after 1 year of treatment,¹⁸ after the second year of treatment the relaxation induced by acetylcholine in vessels from the patients treated with cilazapril was no longer different from that of normotensive subjects. Acetylcholine-induced relaxation of arteries from atenolol-treated patients, in contrast, remained unchanged with treatment even after 2 years of effective blood pressure control.

Discussion

The relationship of functional alterations of blood vessels to the pathogenesis of elevated blood pressure is unclear. Enhanced vasoconstrictor response is one of the features often referred to in relation to the mechanisms of elevated blood pressure. Another functional change often described is an impairment of endothelium-dependent vascular relaxation. It appears, however, that the first may not be present often, and that rather than there being exaggerated vasoconstrictor responses in hypertensive animals and humans, particularly at the level of small arteries, vascular reactivity is often attenuated.¹⁸²⁰ The exception to this may be the response induced by angiotensin II.²¹ The mechanism for this is variable, but changes in the phenotype and therefore in the physicochemical and biological features of smooth muscle cells in the vascular wall probably play a role. In some cases, as for endothelin-1, perhaps other mechanisms may influence responsiveness, for example downregulation of receptors on smooth muscle cells in response to enhanced vascular production of endothelin-1.²² Although the latter has been demonstrated in some experimental models of hypertension, such as the deoxycorticosterone acetate–salt hypertensive rat, the mechanism underlying the attenuated vasoconstrictor responses of small arteries to endothelin-1 in hypertensive patients¹⁵ remains unknown. In hypertensive rats, blunted endothelin-1 responses of mesenteric small arteries were normalized by treatment with the angiotensin-I–converting enzyme inhibitor cilazapril.²³ This occurred together with correction of the structure of small arteries in these rats. Similarly, concomitant with correction of small artery structure,¹⁸¹⁹ maximum responses to endothelin-1 were corrected to normal by treatment with cilazapril after 1 year¹⁸ and, in the present study, remained not different from those of normotensive subjects after 2 years on cilazapril. The apparently paradoxical enhancement of the responsiveness of small arteries to vasoconstrictors, particularly to endothelin-1, under treatment with cilazapril, which occurred together with normalization of the arterial structure, may represent a nonselective correction of the abnormalities of contractility of vascular smooth muscle present in hypertensive humans,⁴⁵¹⁵ a consequence of a regression toward normal of the hypertensive smooth muscle phenotype. Reduced expression of endothelin-1 in the vascular wall as a result of treatment with an angiotensin I–converting enzyme inhibitor, with upregulation of endothelin receptors in smooth muscle cells, could be another mechanism resulting in improved contractility in response to endothelin-1, but this remains to be proven. Enhancement of vascular responses when the blood pressure is normalized suggests that exaggeration of responsiveness to vasoconstrictors should be considered with caution as an indication of a pathogenic role in elevated blood pressure.

Endothelium-dependent relaxation has been shown by most studies to be impaired in hypertensive animals.²⁴²⁵ In hypertensive humans, acetylcholine-induced relaxation investigated by examining effects on forearm blood flow was attenuated in some studies¹⁶²⁶ but not in others.¹⁷ Responses of subcutaneous small arteries from hypertensive patients to acetylcholine studied in vitro were only slightly impaired¹⁸ or normal.²⁷ The differences between these studies may be the result of greater or less severity of hypertension, investigation of different subsets of hypertensive patients, or other factors not currently well understood. In the present study, the slightly attenuated endothelium-dependent relaxation elicited in vitro by acetylcholine in small arteries showed an improvement in patients in the cilazapril group, as demonstrated by the disappearance of statistically significant differences with the relaxation induced by acetylcholine in vessels from normotensive subjects. This result from the present study agreed with that seen in the study done after 1 year of treatment,¹⁸ and together the results of these two studies suggest that impaired endothelial function does improve under treatment with cilazapril, even if this improvement is marginal. Because the impairment of acetylcholine-induced relaxation is relatively minor in this group of patients, it is not surprising that it is difficult to unambiguously demonstrate significant changes. The data from the first and second year complement each other, and demonstrate the reproducibility and persistence in time of this improvement in endothelial function. In contrast, the arteries of the atenolol-treated patients exhibited no change in the impaired endothelium-dependent relaxation after 2 years of treatment, as was seen after 1 year of therapy.¹⁸ The improvement in endothelial function in the hypertensive population examined under treatment with cilazapril is not unexpected, because amelioration of endothelial function after treatment with angiotensin I–converting enzyme inhibitors has already been demonstrated in studies of experimental hypertensive animals.²⁸ Although the response to nitroprusside or another endothelium-independent vasorelaxant was not tested in this study, previous work failed to demonstrate any impairment of endothelium-independent vasorelaxation in hypertensive patients.¹⁶¹⁷²⁶ The mechanism whereby angiotensin I–converting enzyme inhibitors may improve endothelial function (by reduction of generation of angiotensin II, increased local concentrations of bradykinin, or other mechanisms) remains to be established.

In conclusion, correction of structural abnormalities of small arteries of patients with mild essential hypertension treated for 1 or 2 years with the angiotensin I–converting enzyme inhibitor cilazapril is associated with a normalization of altered endothelium-related vasoconstriction (endothelin-1–mediated) and slight improvement of vasorelaxant function (acetylcholine-induced nitric oxide– or other endothelium-derived relaxant factor–mediated). None of these effects are found in a parallel group of age-, sex-, and weight-matched patients treated with the β-blocker atenolol. These results suggest that treatment with angiotensin I–converting enzyme inhibitors may improve both smooth muscle phenotype and impaired endothelial function in patients with mild essential hypertension. Whether these changes in small artery function under antihypertensive treatment translate into improved control of blood pressure and reduced morbidity or mortality is an important question that remains to be answered.

Reprint requests to Ernesto L. Schiffrin, MD, PhD, Clinical Research Institute of Montreal, 110, Pine Avenue W, Montreal, Quebec, Canada H2W 1R7.

• Download figure
• Download PowerPoint