Aortic Stiffness and Left Ventricular Mass in a Rat Model of Isolated Systolic Hypertension
Abstract We tested whether cardiac mass can be related to decreased aortic stiffness in an original rat model of isolated systolic hypertension. Increased aortic stiffness was produced by calcium overload of elastic arteries after vitamin D3 plus nicotine treatment. Half of the animals were chronically treated with the angiotensin-converting enzyme inhibitor perindopril (1 mg/kg per day PO). Rats were pithed, and lower body vascular resistance was measured. Blood pressure was then increased by phenylephrine infusion, and carotidofemoral pulse wave velocity was measured. This value together with those for thoracic aorta internal diameter and medial thickness (determined after in situ fixation and histomorphometry) were used to calculate elastic modulus. Vitamin D3 plus nicotine treatment produced parallel increases in cardiac mass and elastic modulus, with a significant correlation between the two. There was no significant change in resistance. Treatment with perindopril reversed the changes in cardiac mass and elastic modulus but had no effect on resistance after calcium overload of the elastic arteries. In this model of isolated systolic hypertension, we showed that cardiac mass is related to arterial elasticity.
In elderly patients suffering from isolated systolic hypertension, cardiac mass is related not only to total vascular resistance but also to aortic stiffness.1 Thus, the left ventricular mass-volume ratio is significantly correlated with carotidofemoral pulse wave velocity (but not with diastolic arterial blood pressure2 ). The idea has been put forward that (1) increases in impedance and wave reflection amplitude after a chronic increase in aortic stiffness have adverse effects on cardiac function and mass, and (2) these changes can be reversed by a drug-induced decrease in stiffness. The hypothesis is attractive in that it opens up new horizons for antihypertensive therapy,1 3 but there is scarce direct experimental support for the hypothesis. Data for humans are contradictory,4 and animal models for chronic experiments are lacking. Some acute data are available but are mostly negative. Thus, for example, in the ganglion-blocked, epinephrine-maintained normotensive dog (a model similar to the pithed, phenylephrine-maintained normotensive rat that we used; see below), bypass of blood from the aorta to a stiff Tygon conduit produced a dramatic increase in several measures of pulsatile arterial load with little change in cardiac function.5 The situation may be completely different if the heart continues to pump for several weeks or months into a stiff arterial system. This study represents an attempt to provide data from a chronic experiment on the relative importance of central arterial distensibility versus peripheral resistance on cardiac mass.
We used a rat model of isolated systolic hypertension produced by treatment with vitamin D3 and nicotine (VDN). Several indicators (characteristic impedance, systemic arterial compliance, in situ carotid compliance, and pulse wave velocity) reveal a decrease in distensibility after calcium overload of the compliance arteries produced by hypervitaminosis D plus nicotine in this VDN model.6 However, there are no changes in either total vascular resistance or mean arterial blood pressure.6 In chronically implanted, unrestrained VDN rats an increase in systolic and pulse pressures and a decrease in diastolic pressure but no change in mean arterial blood pressure were observed.7 This profile of isolated systolic hypertension together with increased aortic stiffness has been seen in pentobarbital-anesthetized VDN rats in five separate experiments.6 8 9 10 11 The first objective of the present study was to test whether increased aortic stiffness (estimated from carotidofemoral pulse wave velocity and thoracic aorta elastic modulus) could lead to an increase in left ventricular mass in this VDN model.
VDN treatment was initially proposed as an animal model of age-linked calcium overload of compliance arteries in humans.12 13 Vascular calcium overload may be involved in other features of aging of compliance arteries such as dilatation and increased stiffness.14 Calcium fixes preferentially on elastic fibers,8 15 16 and calcification of the vessel wall will produce fracture of elastic fibers and dilation17 and an increase in stiffness.18 Therefore, the second objective of this study was to determine whether calcium overload is linked to increased aortic stiffness in this VDN model.
As stated above, it has been suggested that a drug-induced decrease in aortic stiffness can lead to a reversal of cardiac hypertrophy. Many studies have shown that treatment with angiotensin-converting enzyme inhibitors (ACEI) will reduce arterial stiffness and reverse cardiac hypertrophy.19 20 21 It is uncertain whether the ACEI-induced reduction in arterial stiffness (possibly via an “anticalcinotic” effect12 ) causes reversal of cardiac hypertrophy or whether the changes are parallel but not causal. The third objective of this study was to try to establish whether ACEI-induced reduction in arterial stiffness can lead to a reduction in cardiac mass.
Animals and Chronic Drug Treatment
Ninety-six male, outbred Wistar rats (WI/Ico; body weight, 227±8 g; Iffa-Credo, L’Arbresle, France) were separated into two groups; half received vitamin D3 (300 000 IU/kg IM; cholecalciferol, Wander) and nicotine (2×25 mg/kg PO; bitartrate, Sigma Chemical Co) on day 0 (VDN), and half received saline (0.15 mol/L NaCl, 1 mL/kg IM) and water (2×5 mL/kg PO; control). Subgroups received either the ACEI perindopril in the drinking water or water only (placebo). ACEI treatment was performed from the first to the penultimate day of the experiment (29th or 89th day).
Body weight and food and water intakes were recorded every 3rd or 4th day, and the concentration of perindopril was modified as required in order to maintain a constant dose of 1 mg/kg per day. Spectrophotometric analysis showed that the perindopril solutions were stable for ≤6 days (results not shown). This dose of perindopril was chosen because it does not lower mean arterial blood pressure in VDN rats9 but is hypotensive in normotensive Wistar rats.22 Systolic arterial blood pressure and heart rate were recorded in awake rats before and 30 and 90 days after the onset of treatment using a tail-cuff method in prewarmed animals.20 Hemodynamics was measured in pithed rats on the 30th or 90th day.
Hemodynamics in the Pithed Rat
Twenty-four hours after the final measurement of systolic arterial blood pressure, rats were anesthetized with a halothane (2%)–oxygen mixture. Nylon cannulas (1.02 mm external diameter, 0.58 mm internal diameter; Portex-LSA) were introduced into the right common carotid and left femoral arteries up to the aortic ostia. The anatomic locations of the tips of the cannula were checked by postmortem dissection. The left common carotid artery was tied off. Rats were pithed and artificially ventilated (50 strokes per minute, 12 mL/kg) with air. We used a technique previously described,23 with the following differences: Atropine was not administered, the jugular veins were not tied off, and the vagi were not cut. Blood samples were taken at regular intervals to check that blood gas status remained stable. A full description of this hemodynamic technique has already been published.21 In this section, we will concentrate on the three additional parameters that were measured in this experiment: pulse wave velocity, aortic blood velocity at varying mean arterial blood pressures, and elastic modulus.
Cannulas were connected to low-volume pressure transducers linked to a MacLab-MacBridge analog-to-digital convertor plus computer (AD Instruments Ltd). The frequency response of the cannula plus pressure transducer was flat within ±5% up to 25 Hz; at 25 Hz the phase lag was −7°. The energy of the moduli above the 4th harmonic (<20 Hz in the pithed rat) is similar to the background noise of the recording system (unpublished results).
Systolic, diastolic, mean, and pulse arterial blood pressures (millimeters of mercury) were recorded. Values are given for the right common carotid artery–ascending aorta signal; results obtained at the left femoral artery–iliac bifurcation for mean arterial blood pressure were similar. In the control/placebo group, the average value for the slope relating carotid mean arterial blood pressure to femoral mean arterial blood pressure was 0.97±0.05 mm Hg/mm Hg (n=12 rats, n=35 observations per rat). In the other seven groups, slopes were similar and never significantly different from 1. Heart rate (beats per minute) was calculated from the time lag between successive feet of the carotid arterial blood pressure wave.
Aortic pressure pulse wave velocity was measured by determining the delay(s) between the feet of each systolic arterial blood pressure wave front at the carotid and femoral levels. The computer program defined the foot as the point obtained by extrapolating the diastolic arterial blood pressure decay and the following systolic upswing down to their point of intersection (see Reference 2424 , page 233). The rapidly rising front of the systolic pressure wave has relatively high-frequency components, so we expected measurement of the velocity of the wave front by this simple foot-to-foot method to provide a good estimate of the true phase velocity in the aorta determined by harmonic analysis. In the pithed rat, in which the fundamental harmonic is high (4 to 5 Hz), foot-to-foot pulse wave velocity is more accurate and simpler to measure than other indexes such as characteristic impedance. The distance (in millimeters) between the two cannula tips was determined by direct measurement after in situ fixation (see below) and dissection of the aortic pathway. A damp cotton thread was “stuck” onto the aorta between the tips of the two cannulas, which were marked on the thread. The thread was removed and laid straight for measurement of the distance between the two marks (in millimeters).
Reproducibility of the method was determined in male, outbred Wistar rats (n=5; weight, 510±10 g; age, 12 months). For each rat, the coefficient of variability of repeated measurements of wave velocity before phenylephrine infusion and at stable normotension after phenylephrine infusion was calculated. Values were similar in all rats (global means, 1.7±0.5% and 1.2±0.1% before phenylephrine infusion and at stable normotension, respectively).
Upper abdominal aortic blood velocity was measured using a pulsed Doppler flowmeter (20-MHz high-velocity pulsed Doppler HVPD-20, VF-1, probe HDP-20-XX-S; diameter, 2 mm; Crystal Biotech; frequency response flat up to 30 Hz). The probe was placed around the abdominal aorta immediately below the diaphragm downstream of the celiac artery. Measurement of upper abdominal aortic blood velocity was performed as thoracotomy−required for the measurement of ascending aortic blood flow−produced an extremely low and unstable blood pressure in pithed rats25 and limited the rise in systolic arterial blood pressure produced by phenylephrine to values ≤90 mm Hg (unpublished results).
Because the circulation in both carotid arteries was stopped, we expected upper abdominal aortic blood velocity to provide good approximation of cardiac output (minus coronary blood flow). To check this, thoracotomy was performed in a preliminary experiment in male, outbred Wistar rats (n=4; body weight, 516±24 g; age, 12 months). An electromagnetic flow probe was placed over the ascending aorta (probe diameter, 2.5 mm; flowmeter RT-500, Narco Biosystems). A pulsed Doppler flow probe was placed around the upper abdominal aorta as described above. Seventy measurements were made on each animal over flow ranges of 0 to 45 mL/min. Results were similar in all four rats and were pooled. There was a linear relation between the two signals (F=313, P=.0001). Coefficients for the regression equation Doppler Shift (kHz)=a+(b×[mL/min]) were r=.733, a=0.382±0.155, P for t test for a≠0: P>.05, b=0.103±0.006, P for t test for b≠0: P<.05.
Thirty minutes after implantation of the pulsed Doppler flow probe, lower body peripheral resistance (mean arterial blood pressure divided by lower body blood flow, millimeters of mercury per milliliter per minute) was measured. Because the pithed rat is in a state of near-maximal dilatation,26 such a resistance value is probably close to that measured in the rat hindquarters preparation after maximal dilatation27 and should provide an indirect indication of the nonautonomic mediated diameter of the resistance vessels. As stated above, upper abdominal aortic blood velocity should provide a good approximation of cardiac output, so lower body vascular resistance should be a good indicator of total vascular resistance.
Systolic arterial blood pressure (femoral artery) was then restored back to the level previously recorded for that particular animal (using the tail-cuff method, 24 hours before pithing) by infusion of the α1-adrenoceptor agonist and arteriolar vasoconstrictor phenylephrine, as previously described.21 28 The rationale of this methodology is based on the following assertions. First, an increase in intraluminal distending pressure produces an increase in pulse wave velocity as the elastic modulus of the aortic wall increases after stretching (Reference 2424 , pages 95 through 101). Second, pulse wave velocity is linearly related to mean pressure at values ≤160 mm Hg both in vitro29 and in vivo.21 Third, for evaluation of pressure-independent changes in wall mechanical properties, comparisons should be made under isobaric conditions.30 We propose that a comparison of the slopes relating pulse wave velocity to mean arterial blood pressure provides information on pressure-independent changes in wall stiffness.
Because we wished to study the impact of elastocalcinosis on vessel stiffness, pulse wave velocity was measured over a pressure range of 30 mm Hg up to normotension. At such pressures, transmural strain is borne by elastin (Reference 2424 , page 87).
Pulse wave velocity, lower body blood flow and peripheral resistance, and mean arterial blood pressure were recorded at 1- to 2-minute intervals throughout the 30-minute phenylephrine infusion. Linear regression ANOVA of hemodynamic parameters versus mean arterial blood pressure (independent variable) was performed (N≥15 for each rat). Fig 1⇓ shows the results for pulse wave velocity in a representative of the 30-day control/placebo group. Each individual value for these and the previous hemodynamic measurements used in this and other statistical analyses was the mean of the values obtained over ≥15 heart cycles, that is, ≥3 respiratory cycles. After linear regression ANOVA for each rat, values for intercepts, slopes, and probabilities were averaged (n=12 rats per group).
As stated above, phenylephrine infusion (dose range, 40 to 80 nmol/kg per minute) was adjusted so as to raise femoral systolic arterial blood pressure to a value equal to systolic arterial blood pressure recorded by the tail-cuff method. Pulse wave velocity and mean arterial blood pressure were recorded. The latter was taken as the pressure at which in situ fixation of the aorta was performed for histomorphometric analysis (see below). Arterial dimensions (wall thickness and diameter) and the value for pulse wave velocity were determined at the same level of mean arterial blood pressure. These values were used to estimate the elastic modulus (N/cm2).31
Plasma Renin Activity and Angiotensin-Converting Enzyme Activity
At the end of the experiment, a blood sample (3 mL) was taken from the carotid artery cannula for the determination of plasma renin activity32 and plasma angiotensin-converting enzyme (ACE) activity.33
Aortic Structure, Calcium Content, and Left Ventricular Mass
After blood sampling, rats were perfused for 1 hour at a mean pressure level determined as described above (see “Hemodynamics in the Pithed Rat”) via the right carotid arterial cannula with buffered formol. This histomorphometric technique has been described in detail previously.21 After in situ fixation, the thoracic aorta was excised, immersed in formol, dehydrated, and embedded in paraffin. Sections stained with hematoxylin plus eosin were used for the determination of inner diameter, medial thickness, and the ratio of medial thickness to internal diameter. Sections stained with Weigert’s solution were used for the determination of the percentage of the medial surface occupied by elastin. Morphometric analysis was performed using the Optilab algorithm (Graphtek, 93510). Each section was examined three times in a blind fashion.
A sample of the thoracic aorta was removed for determination of calcium (micromoles per gram tissue dry weight) and protein (milligrams per gram tissue wet weight) contents.8 The arterial calcium content was compared with that found in other soft tissues (myocardium, kidney, and small intestine). The heart was removed, and the left ventricle was dissected out and weighed (milligrams). Because treatment with ACEI produced a significant fall in body weight (see “Results”), left ventricular mass was expressed relative to body weight (milligrams per kilogram). After weighing, a sample of the left ventricle was removed for the determination of calcium content (see above).
Randomization was carried out on the basis of the initial (compared with before treatment with either VDN or ACEI) values for systolic arterial blood pressure. A balanced, randomized block design was used with three-way ANOVA of the three independent factors: VDN, ACEI, and time. Variability related to the three-way interaction was incorporated into the error mean square. For any given parameter, missing values per group were less than or equal to two. The Bonferroni test was used for the comparison of means, which are given as ±SEM (n=number of rats, N=number of observations per rat). Regression ANOVAs were performed using standard parametric techniques; results are expressed as intercept (a) and slope (b). In some experiments, individual values for intercepts and slopes obtained in any one given preparation with N observations were treated as individual, continuous variables to which standard parametric analyses were applied.
Experiments followed the guidelines of the European Union, the French Ministry of Agriculture, and the University of Nancy concerning experimentation on live mammals.
Body Weight, Food and Water Intakes, Systolic Arterial Blood Pressure, Heart Rate, ACE, and Plasma Renin Activities
VDN rats initially ate less and lost weight, then recovered. By the 30th day, they had a body weight 94% that of control rats; at 90 days, body weight was the same as in controls. ACEI produced a significant, 10% decrease in body weight in control rats at 90 days. Food and water intakes were similar in all groups at the time of hemodynamic experiments (see Table 1⇓).
VDN treatment produced an increase in systolic arterial blood pressure of 12% at 30 days and 18% at 90 days. ACEI lowered systolic arterial blood pressure at 30 days (−14% and −18%) and 90 days (−23% and −30% in control and VDN rats, respectively). The ACEI×VDN interaction was significant, suggesting a significantly greater effect of ACEI in VDN rats. Heart rate was unaffected by VDN, ACEI, or time (control/placebo: 438±15 beats per minute at the start of the experiment).
ACE activity rose and plasma renin activity fell between 30 and 90 days. ACEI-induced falls in ACE activity and increases in plasma renin activity were similar in control and VDN rats.
Left Ventricular Mass and Aortic Length
There was no significant difference in the slopes relating left ventricular mass to body weight between control and VDN rats (1806±246 and 1430±330 mg/kg body wt, respectively). In the placebo groups, left ventricular mass was 16% greater in VDN than in control rats (Fig 2⇓). ACEI reduced mass in controls by −15% and −11% and in VDN rats by −16% and −22% at 30 and 90 days, respectively.
There was a significant linear regression between aortic length and body weight (a, 107±5 mm; b, 76±12 mm/kg; n=96). Neither VDN nor ACEI had any effect on the slope relating aortic length to body weight (results not shown) or on length of the aorta. Aortic length increased globally with time: 30 days, 136±10; and 90 days, 143±10 mm (P<.05).
Histomorphometry of the Thoracic Aorta
VDN treatment produced a significant increase in diameter of +20% at 90 days. ACEI produced a significant decrease in internal diameter in VDN rats. VDN had no effect on medial thickness. ACEI produced a significant decrease in medial thickness and in the ratio of medial thickness and internal diameter in controls. VDN produced a significant 10% reduction in the percentage of the medial surface occupied by elastin at 90 days; ACEI counteracted this reduction (see Table 2⇓).
Calcium Content of Thoracic Aorta
VDN treatment produced a 30-fold increase in the calcium content of the thoracic aorta that was counteracted by 90 days of ACEI treatment (Fig 3⇓). VDN treatment also produced an increase in the calcium content of the myocardium (control, 3.5±0.8; VDN, 27.7±3.9 μmol/g; P<.05) and the kidney (control, 6.5±0.6; VDN, 86.6±13.0 μmol/g; P<.05). ACEI had no effect. The intestinal calcium content (overall mean, 11.9±0.7 [SD] μmol/g) was similar in all groups. Acid-soluble protein contents were similar; global means (±SD) were thoracic aorta, 18±9 mg/g wet wt; myocardium, 16±9; kidney, 37±3; and intestine, 22±8.
Hemodynamic Parameters in Pithed Rats
VDN treatment had no effect on baseline lower body peripheral resistance (Fig 4⇓). ACEI produced a significant fall in resistance of 50% in controls.
In most cases the slopes relating lower body blood flow to mean arterial blood pressure during phenylephrine infusion were not significant (Table 3⇓). Lower body peripheral resistance was closely correlated to mean arterial blood pressure (Table 3⇓). Neither VDN nor ACEI had any effect on the slope relating resistance to pressure. Care should be taken, however, when interpreting the results on resistance versus pressure, as the latter is the numerator of the former.
Pulse wave velocity was closely related to mean arterial blood pressure. Values for slopes were similar in all controls. VDN produced a 71% increase in slope. Ninety but not 30 days of ACEI treatment produced a normalization of the slope in VDN.
VDN produced a twofold increase in calculated aortic elastic modulus (Fig 5⇓). ACEI reduced elastic modulus in VDN by more than 50% but had no effect on elastic modulus in control rats.
Correlations Between Cardiovascular Structure and Hemodynamic Parameters in VDN Rats
Linear regression ANOVA was performed on the data from the two age groups (30 or 90 days) of the control/placebo or VDN/placebo rats with the following parameters: (1) dependent variable: elastic modulus; independent variables: thoracic aorta calcium content or internal diameter, (2) dependent variable: velocity/pressure slope; independent variable: thoracic aorta calcium content or internal diameter, and (3) dependent variable: left ventricular mass; independent variables: lower body peripheral resistance, velocity/pressure slope, elastic modulus, or thoracic aorta internal diameter.
A significant regression was found between left ventricular mass (mg/kg body wt, Y) and elastic modulus (N/cm2) in the 30-day VDN/placebo group: a=1544±219 (P<.05), b=13.3±4.8 (P<.05). Applying this equation to the value for elastic modulus obtained in the VDN rats treated for 30 days with ACEI (17±1 N/cm2), a calculated left ventricular mass of 1770 mg/kg was found. This was very close (0.7% less) to the value (1783±51 mg/kg) actually measured in this group. In the 90-day VDN/placebo group the regression equation was similar: a=1588±225 (P<.05), b=9.9±3.6 (P=.08), but the slope failed to reach significance.
At 30 (but not 90) days, both elastic modulus (F=4.5, P<.05) and left ventricular mass (F=13.3, P<.05) were significantly correlated with thoracic aorta internal diameter. Regarding the first correlation, it should be noted that internal diameter enters into the Bergel equation for elastic modulus.
Because this study uses a new model for the measurement of vascular mechanics, the pithed rat, several methodological points will be dealt with first. The pithed rat has been used extensively to study arteriolar vasoconstrictor agents, to which it has a very high sensitivity.23 We have previously used the pithed rat to investigate ACEI-induced changes in aortic stiffness by using the velocity/pressure slope21 ; the technique used in the present experiment is based on our earlier study. This is to our knowledge the first description of the use of pithed rat for the measurement of elastic modulus.
The pithed rat has the advantage that arterial stiffness can be measured over a wide mean arterial blood pressure range of approximately 80 mm Hg, twice the range in previous studies.34 Furthermore, in the previous studies in anesthetized rats,34 blood pressure varied spontaneously with (presumably) a bell-shaped distribution about a mean value. This would mean that the compliance values determined in the center of the blood pressure range would be based on a far greater number of values than those obtained at the two extremes of the blood pressure range. In our experiments, the number of measurements of pulse wave velocity was approximately evenly distributed over the blood pressure range (Fig 1⇑). In the pithed rat, motion artifacts, which can interfere with hemodynamic measurements using techniques such as echo-tracking, etc, are absent (apart from respiration).
Pulse wave velocity was measured over the whole length of the aorta, whereas geometry was determined at the thoracic level only. This probably does not invalidate our technique because the values for elastic modulus in control/placebo rats are comparable to those previously reported for the rat aorta (Reference 2424 , page 76). Another source of potential error is that phenylephrine may have a direct effect on aortic stiffness independent of its indirect effect via a change in transmural pressure after an α1-adrenoceptor–mediated increase in arteriolar resistance. This seems unlikely, however, because in male, outbred Wistar rats (n=6; weight, 606±8 g; age, 12 months) in which resistance was increased by perfusion of phenylephrine or angiotensin II, values for the velocity/pressure slopes were similar (phenylephrine, 54±4; angiotensin II, 60±2 [mm/s]/mm Hg).
In our VDN model of isolated systolic hypertension, elastic modulus and left ventricular mass increased in parallel. At 30 days, there was a significant linear correlation between the two, and the regression equation obtained in VDN/placebo accurately predicted the left ventricular mass in VDN rats after ACEI treatment. At 90 days, the regression was similar but failed to reach statistical significance. Although correlations should always be interpreted with caution, it should be noted that despite the fact that the number of animals per group was not large, a significant correlation was obtained using the data of a single group and not (as is sometimes the case) by mixing the data of two groups (one treated and one control) and obtaining a regression line between two isolated clusters of data points. Cardiac mass in VDN was not related to (lower body) vascular resistance. We have previously reported that there are no significant changes in total vascular resistance for up to more than 1 year after VDN treatment.6 The results obtained with lower body vascular resistance in the present study confirm this.
To our knowledge, this is the first description of a link between central arterial stiffness and left ventricular mass in a “normotensive” rat model. It should be noted that while we have extensively validated the VDN rat as a normotensive model of vascular calcium overload with increased impedance and pulse wave velocity, reduced compliance and isolated systolic hypertension,6 7 8 9 10 11 and increased elastic modulus (present results), it is obvious that this animal model reproduces only one facet of the extremely complex mosaic of isolated systolic hypertension in the elderly, that is, vascular calcium overload leading to increased arterial stiffness.
ACEI treatment in VDN rats produced parallel falls in elastic modulus and left ventricular mass but had no significant effect on vascular resistance. This observation agrees with our hypothesis that in the VDN model of isolated systolic hypertension, cardiac mass is linked to central arterial stiffness but not to vascular resistance. The opposite is observed with ACEI treatment in control rats, in which parallel changes in cardiac mass and vascular resistance with no change in elastic modulus were observed. The link in controls between ACEI-induced changes in cardiac mass and vascular resistance confirms recently published results35 obtained by treating normotensive Wistar rats with the same perindopril regimen as used in the present study.
These observations raise the possibility that an ACEI-induced fall in left ventricular mass may pass via a mechanism dependent on a fall in elastic modulus only in those cases in which elastic modulus was initially abnormally elevated. This could be the case for ACEI treatment of isolated systolic hypertension in humans. It also could explain why the ACEI-induced degree of change in arterial stiffness was not significantly correlated to the ACEI-induced reduction in left ventricular hypertrophy in a recent study on patients with essential hypertension and increased (rather than decreased) diastolic pressure.4
If we now consider how ACEI could produce a change in elastic modulus, aortic geometry will be dealt with first. In control rats, ACEI produced a decrease in thoracic aorta medial thickness but had no effect on internal diameter. This pattern is the same as that observed by Michel et al35 after chronic ACEI (same perindopril regimen as the present study) in normotensive Wistar rats. We would suggest that in such animals, the ACEI-induced fall in vascular resistance (total for Michel et al, lower body in present study) leads to a fall in mean arterial blood pressure (see Michel et al and Lartaud et al22 for the same perindopril regimen in normotensive Wistar rats, not measured in this study), which in turn leads to a decrease in aortic medial thickness. In VDN rats, the pattern is entirely different. ACEI has no effect on aortic medial thickness but counteracts the increase in internal diameter. Such changes occur in the absence of any modification of vascular resistance in VDN rats, suggesting that contrary to controls, changes in aortic geometry in VDN rats are not the indirect consequence of changes in blood pressure.
We suggest that ACEI modifies the process of elastocalcinosis, leading to dilation and increased stiffness described in the introduction. After 30 days, the ACEI effect is witnessed by a decrease in aortic internal diameter. It is more obvious after 90 days, as in addition to a fall in aortic internal diameter, there is an increase in elastin (possibly a protection against elastin degradation after calcium deposition) and a decrease in calcium content of the aortic wall. The change in the composition of the aortic wall after 90 days of ACEI presumably explains why the wave velocity/pressure slope, an index of pressure-independent changes in stiffness, is normalized after 90 (but not after 30) days of ACEI. The exact mechanisms by which, in the VDN model, an increase in elastic modulus leads to an increase in cardiac mass and ACEI reverses this pattern are unknown.
One weakness of our study must be raised, namely, that central aortic blood pressure was not measured. Thus, we cannot eliminate the possibility that in VDN rats either (1) ACEI treatment lowered mean arterial blood pressure and hence transmural pressure, leading to a reduction in the elastic modulus of the aortic wall, which in turn produced a reduction in left ventricular mass, or (2) ACEI treatment lowered mean arterial pressure, leading to parallel but unrelated changes in elastic modulus and left ventricular mass. However, we have previously shown that the dose of perindopril we used did not lower mean arterial blood pressure in VDN rats,9 and in the present study, ACEI did not diminish (lower body) vascular resistance. It should be noted that in this discussion, no argument is developed on the basis of tail systolic arterial blood pressure because the relation between this pressure and aortic pressure could be modified by changes in wave reflection transmission by the VDN and/or ACEI treatments (see Ting et al36 and specifically for perindopril, London et al37 ). If wave reflection transmission changes then, the time at which the reflected wave merges with the systolic part of the pulse would change, and this would modify central systolic pressure. To investigate this possibility we measured the augmentation index (percent) and the left ventricular ejection time/travel time of the reflected wave (LVET/Δtp) of the carotid artery pressure waveform according to previously described methods37 38 in awake normotensive rats, normotensive rats treated with perindopril (1 mg/kg per day PO) and VDN rats (n=5 per group) of approximately the same age as the rats used in this study. Augmentation indexes and LVET/Δtp values were 25±2 and 4.1±0.4, 27±3 and 3.5±0.4, and 25±1 and 3.0±0.4 in normotensive rats, normotensive rats treated with perindopril, and VDN rats, respectively. Fourier analysis of the carotid artery waveform was performed, and no differences between the values for the first six harmonics among the three groups were observed. In all three groups the pressure waveform was of the A type described by Murgo et al,38 and this may explain why in the rat values for tail (tail-cuff) and carotid (invasive) systolic arterial pressures measured simultaneously are almost identical.39
Using a new model of isolated systolic hypertension, we have provided evidence for our working hypothesis that cardiac mass is related to aortic stiffness. Further experiments, especially those involving measurement of central aortic pressure in conscious animals, are now required to confirm this working hypothesis.
The authors wish to thank I.R.I.S., 92415 Courbevoie Cedex, France, for financial support. They especially express their gratitude to Dr Nancy Thibault, of the cardiovascular division of I.R.I.S., Courbevoie, for her intellectual, material, and moral support.
Asmar RG, Pannier B, Santoni JP, Laurent S, London, GM, Levy BI, Safar ME. Reversion of cardiac hypertrophy and reduced arterial compliance after converting enzyme inhibition in essential hypertension. Circulation. 1988;78:941-950.
Schartl M, Bocksch WG, Preysse S, Beckmann S, Franke O, Hunten U. Remodeling of the myocardium and arteries by chronic angiotensin converting enzyme inhibition in hypertensive patients. J Hypertens. 1994;12(suppl 4):S37-S42.
Kelly RP, Tunin R, Kass DA. Effect of reduced aortic compliance on cardiac efficiency and contractile function of in situ canine left ventricle. Circ Res. 1992;71:490-502.
Atkinson J, Poitevin P, Chillon JM, Lartaud I, Levy B. Vascular calcium overload produced by vitamin D3 plus nicotine treatment diminishes arterial distensibility in rats. Am J Physiol. 1994;266:H540-H547.
Henrion D, Chillon JM, Capdeville-Atkinson C, Vinceneux-Feugier M, Atkinson J. Chronic treatment with the angiotensin I converting enzyme inhibitor, perindopril, protects in vitro carbachol-induced vasorelaxation in a rat model of vascular calcium overload. Br J Pharmacol. 1991;104:966-972.
Henrion D, Chillon JM, Capdeville-Atkinson C, Atkinson J. Effect of chronic treatment with the calcium entry blocker, isradipine, on vascular calcium overload produced by vitamin D3 and nicotine in rats. J Pharmacol Exp Ther. 1992;260:1-8.
Fleckenstein A, Frey M, Zorn J, Fleckenstein-Grun G. Calcium, a neglected factor in hypertension and arteriosclerosis. In: Laragh JH, Brenner BM, eds. Hypertension: Pathophysiology, Diagnosis and Management. New York, NY: Raven Press; 1990:471-509.
Atkinson J. Vascular calcium overload: physiological and pharmacological consequences. Drugs. 1992;44(suppl 1):111-118.
Nichols WW, O’Rourke MF. McDonald’s Blood Flow in Arteries. London, UK: Edward Arnold; 1990:398-409.
Gertz SD, Kurgan A, Eisenberg D. Aneurysm of the rabbit common carotid artery induced by periarterial application of calcium chloride in vivo. J Clin Invest. 1988;81:649-656.
Dzau VJ, Safar ME. Large conduit arteries in hypertension: role of the vascular renin-angiotensin system. Circulation. 1988;77:947-954.
Lartaud I, Makki T, Bray-Des-Boscs L, Niederhoffer N, Atkinson J, Corman B, Capdeville-Atkinson C. Effect of Ang I-converting enzyme inhibition on aging processes, IV: cerebral blood flow regulation. Am J Physiol. 1994;267:R687-R694.
Shipley RE, Tilden JH. A pithed rat preparation suitable for assaying pressor substances. Proc Soc Exp Biol. 1947;64:453-455.
Milnor WR. Hemodynamics. 2nd ed. Baltimore, Md: Williams & Wilkins; 1989.
Richer C, Lefevre-Borg F, Lechaire J, Gomeni C, Gomeni R, Giudicelli JF, Cavero I. Systemic and regional hemodynamic characterization of alpha-1 and alpha-2 adrenoceptor agonists in pithed rats. J Pharmacol Exp Ther. 1987;240:944-952.
Sano T, Tarazi RC. Differential structural responses of small resistance vessels to antihypertensive therapy. Circulation. 1987;75:618-626.
Vargas HM, Cuevas JM, Ignarro LJ, Chaudhuri G. Comparison of the inhibitory potencies of NG-methyl- NG-nitro- and NG-amino-l-arginine on EDRF function in the rat: evidence for continuous basal EDRF release. J Pharmacol Exp Ther. 1991;257:1208-1215.
Mulvany MJ. A reduced elastic modulus of vascular wall components in hypertension. Hypertension. 1992;20:7-9.
Bergel DH. The dynamic elastic properties of the arterial wall. J Physiol. 1961;156:458-469.
Peters-Haefeli L. Rate of inactivation of endogenous and exogenous renin in normal and renin-depleted rats. Am J Physiol.. 1971;221:1339-1345.
Ryan JW, Chung A, Ammons C, Carlton ML. A simple radioassay for angiotensin-converting enzyme. Biochem J. 1977;167:501-504.
Hayoz D, Tardy Y, Perret F, Waeber B, Meister JJ, Brunner HR. Noninvasive determination of arterial diameter and distensibility by echo-tracking techniques in hypertension. J Hypertens. 1992;10(suppl 5):S95-S100.
Michel JB, Heudes D, Michel O, Poitevin P, Philippe M, Scalbert E, Levy BI, Corman B. Effect of chronic ANG I-converting enzyme inhibition on the aging processes, II: large arteries. Am J Physiol. 1994;267:R124-R135.
Ting CT, Yang TM, Chen JW, Chang MS, Yin FCP. Arterial hemodynamics in human hypertension: effects of angiotensin converting enzyme inhibition. Hypertension. 1993;22:839-846.
London GM, Pannier B, Guerin AP, Marchais SJ, Safar ME, Cuche JL. Cardiac hypertrophy, aortic compliance, peripheral resistance, and wave reflection in end-stage renal disease: comparative effects of ACE inhibition and calcium blockade. Circulation. 1994;90:2786-2796.
Murgo JP, Westerhof N, Giolma JP, Altobelli SA. Aortic input impedance in normal man: relationship to pressure wave forms. Circulation. 1980;62:105-115.
Bunag RD. Validation in awake rats of a tail-cuff method for measuring systolic pressure. J Appl Physiol. 1973;34:279-282.