Acta Physiol Scand 1991, 141, 507-516
ADONIS
0001667729100075G
Right ventricular coronary blood flow patterns during aortic pressure reduction in renal hypertensive dogs J. J. S M O L I C H , P. L. W E I S S B E R G , P. F R I B E R G and P. I. KORNER Baker Medical Research Institute, Melbourne, Victoria, Australia
SMOLICH, J. J., WEISSBERG, P. L., FRIEERG, P. & KORNER, P. I. 1991. Right ventricular coronary blood flow patterns during aortic pressure reduction in renal hypertensive dogs. Acta Ph,ysiol Scund 141, 507-516. Received 17 July 1990, accepted 22 October 1990. ISSN 0001-6772. Baker Medical Research Institute, Melbourne, Victoria, Australia. We measured right ventricular coronary blood flow with radioactive microspheres during graded aortic pressure reduction in 13 normal dogs and in 13 renal hypertensive dogs with left ventricular hypertrophy. Under anaesthesia and controlled loading conditions, mean aortic pressure was lowered from control (128 mmHg in normal and 146 mmHg in hypertensive dogs) to approximately 100, 90 and 80 mmHg. In normal dogs, right ventricular blood flow was not affected by this pressure reduction, consistent with effective right ventricular autoregulation. In hypertensive dogs, however, right ventricular blood flow was maintained between a mean aortic pressure of 146 and 90 mmHg (range 75-79 ml min-' 100 g-') but fell by 18% to 63 ml min 100 g-' at a mean aortic pressure of 80 mmHg ( P < 0.005). We conclude that autoregulation of right ventricular blood flow was preserved in chronic hypertension but that, compared to normal dogs, the lower limit of autoregulation was reset to a higher pressure level. Moreover, the similarity of right ventricular-to-body weight ratios in the two groups implied that this change was a consequence of hypertension-induced structural changes in the coronary vasculature. Key words ; autoregulation ; coronary circulation ; coronary vascular resistance ; hypertension ; radioactive microspheres.
T h e left ventricle (LV) exposed to chronic systemic hypertension commonly exhibits coronary blood flow abnormalities (Marcus 1983). However, because of the coexistence of hypertension-induced vascular changes and muscle hypertrophy in the left ventricle in this condition, it may be difficult to separate the relative contribution of these two factors to any alterations in LV blood flow patterns. T h i s is less of a problem in the right ventricle (RV) because, while the coronary vasculature of this chamber is exposed to the same elevated arterial pressures as the left ventricle, RV muscle hypertrophy is Correspondence : Dr J. J. Smolich, Centre for Early Human Development, Monash Medical Centre, Clayton Road, Clayton, Victoria, Australia, 3 168.
either absent (Tomanek et al. 1985) or of only minor degree (Wangler et al. 1982, Wicker & Tarazi 1985). Previous coronary flow studies of the right ventricle in chronic hypertension (Wangler e t al. 1982, Tomanek et al. 1985, Wicker & Tarazi 1985), have demonstrated that RV blood flow is preserved under control conditions, even though RV coronary vascular resistance is increased. Little is known, however, about the effects of systemic hypertension on RV blood flow patterns during changes in RV perfusion pressure. I n an earlier study in normal dogs (Smolich et al. 1988a), we observed that RV myocardial perfusion, measured with radioactive microspheres, remained essentially constant during reductions in RV perfusion pressure produced
507
508
3.3. Smolich et al.
b y stepwise lowering of aortic pressure. The aim of the present study was to determine the effect of renovascular hypertension on RV blood flow responses during aortic pressure reduction.
MATERIALS AND METHODS Animals. T h e full study protocol was performed in 26 mongrel dogs. T h e experimental group comprised 13 dogs with renovascular hypertension. T h e 13 dogs in the control group were selected on the basis of a similar range of body weights and were studied concurrently with the hypertensive group. All animals received several weeks of health conditioning prior to any surgical procedure. All experimental procedures were in accord with guidelines set by the National Health and Medical Research Council of Australia. Induction of hypertension. One-kidney renovascular hypertension was induced as previously described (Anderson et al. 1979, Broughton & Korner 1983). Briefly, animals were anaesthetized with i.v. sodium thiopentone, intubated and ventilated with a mixture of oxygen, nitrous oxide and halothane. Indwelling polyvinyl catheters were inserted into the abdominal aorta through a left flank incision. An inflatable Silastic cuff (Hazen Everett, New Jersey, USA) and a continuous wave Doppler flow probe were placed around the left renal artery. The right kidney was then removed through a right flank incision. Postoperatively, animals were given oral antibiotics for 7 days and analgesia as required. Recovery from surgery was complete by 1&14 days. Thereafter, aortic blood pressure was measured with the animals fully conscious and unrestrained but lying quietly on a padded table. A gradual rise in arterial blood pressure over several weeks was produced by progressive inflation of the renal artery cuff. The elevated blood pressure was maintained for at least 6 week prior to the acute experiment. Acute experiment. Normal and hypertensive dogs were anaesthetized with i.v. sodium pentobarbitone (20-30 mg kg bolus followed by a constant infusion of 6 mg kg h). The lungs were ventilated at an endexpiratory pressure of 4 cm H,O (0.4kPa) with oxygen-enriched air. Arterial blood gases were checked frequently (Radiometer ABL1, Copenhagen, Denmark) and ventilation was adjusted to maintain P,,, between 35 and 40 mmHg (4.7-5.3 kPa) and arteriaf p H between 7.3 and 7.4. Base deficits were corrected with i.v. sodium bicarbonate as required. Body temperature was maintained between 36 "C and 38 "C with a heated table and water blanket. T h e surgical preparation of the animals was as previously described (Smolich et al. 1988a, 1988b). The chest was opened in the 5th left interspace. Fluid-filled pressure catheters were inserted into the left atrium, and into the ascending aorta via the
carotid artery. A large cannula was introduced into the left atrium through the appendage and connected to a reservoir filled with a mixture of room-temperature heparinized blood and low molecular weight dextran (10% Rheomacrodex in normal saline). A highfidelity micromanometer (Konigsberg P4.5) was calibrated in a 37 "C water bath against a mercury manometer and then inserted into the left ventricle through the apical dimple in all animals as part of another protocol. A second micromanometer (Konigsberg P4.5) was introduced into the RV cavity through its free wall to measure RV pressure in a limited number of control and hypertensive animals. In these animals a catheter was also inserted into the right atrial cavity and/or the pulmonary trunk to provide a pressure reference for the RV micromanometer. The edges of the pericardial incisions were loosely approximated with interrupted sutures. Lastly, a plastic Ypiece was inserted into the thoracic aorta during several minutes occlusion between cross-clamps. An arteriovenous (AV) fistula was created by connecting the protruding arm of the Y-piece to a jugular venous cannula with wide-bore tubing. Flow through the fistula was regulated with a roller pump. After the completion of surgery, the chest wound was left open but covered with towels to minimize heat loss. Experimental protocol. Autonomic ganglionic transmission was blocked with mecamylamine (2 mg kg-' i.v.) and the cervical vagosympathetic trunks were cut to eliminate reflex changes in cardiac contractility arising from reductions in aortic pressure (Broughton & Korner 1983, Smolich et al. 19886). The heart was paced at 150 beats min-' with a stimulator (Grass SD9) and bipolar left atrial electrode. Arterial and atrial pressures were measured with Statham P23D strain gauge transducers using the mid-thoracic vertebral spines as the zero reference level. Openchest mean left atrial pressure was held at 15 mmHg with the left atrial reservoir. Ventricular, arterial and atrial pressures were displayed on a Devices MI6 paper recorder. I n both the normal and hypertensive dogs, mean aortic pressure (MAP) was reduced in steps with the AV fistula from the control value to approximately 100 (MAPlOO), 90 (MAP90) and 80 mmHg (MAP80). Steady state conditions were maintained for several min at each reduced pressure level. Haemodynamic variables were recorded and myocardial blood flow then measured at the control and reduced pressure levels with 15 ,um radioactive microspheres (Heymann et al. 1977). Isotopes were randomly selected from six available labels (141Ce %r l13Sn "Sr 95Nb, 46Sc; New England Nuclear). Approximately 1 million microspheres were mixed in 2 ml of physiological saline for 10 min by ultrasonification and manual agitation and then injected into the left atrial cavity over 30-40 s with 15 ml of saline. A reference sample was withdrawn at 12 ml/min from an aortic catheter
RV coronary bloodjow in hypertension with a mechanical pump (Harvard 940A, Massachusetts, USA), starting 5-10 s before injection and ending 75-90 s after injection. Cardiac output was measured with the thermodilution technique (Korner & Hilder 1974) in seven normal and seven hypertensive dogs during the pressure reduction. Room-temperature 5 yo dextrose was injected into the right atrium and the resultant time-temperature curves were recorded with a thermistor probe situated in the descending aorta or a major pulmonary artery. Cardiac output was determined in duplicate and indexed for body wt. RV and mean right atrial pressures were also measured in a separate group of three normal animals which underwent aortic pressure reduction according to the above protocol, but which did not receive the full complement of microsphere injections. Fixation of hearts. At the end of the experiment, two normal and two hypertrophied hearts were arrested in diastole with KCI, cleared of blood and fixed at a perfusion pressure of 100 mmHg with 2% paraformaldehyde and 2% glutaraldehyde in 0.1 M phosphate buffer. Segments of the left anterior descending coronary artery were excised from proximal, mid and distal sites. These were fixed for a further 2 h in 1% osmium tetroxide in 0.1 M phosphate buffer, dehydrated in acetone and embedded in eponaraldite. Semi-thin sections were cut with glass knives and stained with 1% methylene blue. The sections were projected onto a computer-linked digitizing tablet and the cross-sectional area of the media and the length of the internal elastic lamina measured. As neither are appreciably influenced by vessel contraction (Lee et al. 1983, Greensmith & Duling 1984), linear regression of medial cross-sectional area on the length of the internal elastic lamina provided a measure of the wall-to-lumen ratio which was not affected by any fixation-related differences in vessel contraction between normal and hypertensive animals (Owens et al. 1988). The other dogs were sacrificed with an overdose of sodium pentobarbitone and potassium chloride at the end of the experiments and the hearts immersed in fixative for 7-10 days. Myocardial blood jlow measurements. The large coronary vessels, epicardial fat and valves were removed. The RV free wall was separated from the septum and dissected into 9 endocardial and 9 epicardial samples (Smolich et al. 1988a). In the normal dogs, the former constituted 40.1 f0.9% and the latter 59.9f0.9y0 of the total RV weight. The corresponding contributions of these two layers in the hypertensive dogs were 39.6 f0.7% and 60.4 0.7% respectively. The right quarter of the interventricular septum, which is anatomically continuous with the RV subendocardium, was also removed and divided into 6 segments. 18
509
Isotope radioactivity in the reference (r) and tissue (t) samples was determined with a Packard 5130 through-hole gamma counter. The radioactivity of each isotope peak in the tissue samples was obtained by stripping away the contribution of background and the Compton areas of any higher energy isotopes with a computer program. Blood flow to the tissue samples was calculated from the relation: Q, = (Qx R,)/R,, where Q is flow (ml min-') and R is radioactivity (counts min-l). Total flow to the RV free wall and the right side of the interventricular septum was obtained by summing the individual sample flows. Flows to the various full-thickness regions, and the endocardial and epicardial halves of the RV free wall were calculated by appropriate combination of the same sample flows. All flows were normalized to 100 g of wet tissue. Coronary vascular resistance. Relative coronary vascular resistance of the RV free wall was calculated where QM,,is average RV from (PA,, - PA,,)/&,,,, blood flow (ml min 100 g-'), PART is mean arterial pressure (kPa) and PA,, is mean right arterial pressure (kPa). Mean right atrial pressure was assumed to be 11 mmHg (1.5 kPa) in control dogs and 9 mmHg (1.2 kPa) in the hypertensive animals (Table 2). Statistical analysis. Results were evaluated with standard statistical tests (Snedecor & Cochran 1980). Changes in haemodynamic variables and RV blood flow in normal and hypertensive dogs at the various pressure levels were analyzed with two and three way analysis of variance (ANOVA). The between pressures and, where appropriate, the between regions sums of squares was orthogonally partitioned into individual degrees of freedom. Differences between normal (N) and hypertensive (HT) dogs were evaluated with t = A(HT - N)/(SEi SE;,) 0.5 where A(HT - N) was the difference between the response of normal and hypertensive dogs and SEN and SE,, were the appropriate standard errors from analysis of variance (Broughton & Korner 1983). The significance of decreases in RV coronary vascular resistance during aortic pressure reduction was assessed with a one tailed paired t-test. Differences in the relationship between medial cross-sectional area and the length of the internal elastic lamina in coronary arteries of normal and hypertensive dogs were analyzed with analysis of covariance, using the pooled regression coefficient of each group. A value of P < 0.05 was considered significant. Unless specified otherwise, results are expressed as mean SE.
+
RESULTS Mean arterial pressure rose from an average baseline value of 103f2 m m H g to 131f4 mm1 ig following renal artery constriction in the experimental group of dogs. T h e elevated blood ACT 141
510
3.3.Smolich et al.
Table 1. Body and ventricular weights in the normal and hypertensive dogs
BW, kg LVW/BW, g/kg RVW/BW, g/kg
Normal ( n = 13)
Hypertension ( n = 13)
27.4f 1.7 4.16k0.16 1.30f0.05
27.3 f 1.2 5.13 f0.23 1.36k0.04
P NS
< 0.005 NS
BW, body weight at time of acute experiment; LVW, left ventricular weight; RVW, right ventricular weight. P, normal us. hypertension, unpaired two tailed t-test ; NS, not significant.
Table 2. Haemodynamic parameters in the control state (C) and three reduced pressure levels (PI, P,, P3) in normal and hypertensive dogs C MAP, mmHg Normal ( n = 13) Hypertension ( n = 13) RVSP, mmHg Normal (n = 6) Hypertension (n = 7) RVDP, mmHg Normal ( n = 6) Hypertension ( n = 7) MRAP, mmHg Normal ( n = 6) Hypertension ( n = 5) CI, ml min kg-' Normal (n = 7) Hypertension ( n = 7)
SE
p3
Pl
____
128 146"*
104 107
91 91
78 77
3 4
31 31
32 30
31 30
31 31
1 1
5
6 6
5 6
5 7
1 1
10.8 9.0
10.9 9.0
11.0 9.3
0.4 0.3
8XXX 10.2" 9.3 149 157
175 179
195 199
202 187
11 9
MAP, mean aortic pressure ; RVSP, right ventricular systolic pressure; RVDP, right ventricular minimal diastolic pressure; MRAP, mean right atrial pressure; CI, cardiac index; SE, SE of the difference between any two means within animals from two-way ANOVA. ** P < 0.025, normal us. hypertension, unpaired t-test. * P < 0.05, *"" P < 0.005, compared to reduced pressure levels, two-way ANOVA.
pressure was maintained for an average of 10 week (range 6-17 week). At postmortem, the LV-to-body weight ratio of the hypertensive dogs was 23 yogreater than in normal dogs (P< O.OOS), but the RV-to-body weight ratios were closely similar in the two groups (Table 1).
Haemodynamic indices During the initial control period under anaesthesia, mean aortic pressure in the hypertensive animals was significantly higher than in the normal dogs, but this variable was matched in the two groups at lower pressures (Table 2). Phasic RV pressures were similar in hypertensive
and normal dogs at all pressures. With an identical mean left atrial pressure, mean right atrial pressure in normal and hypertensive dogs did not differ significantly under control conditions. However, when the grouped values from all aortic pressure levels were compared, mean right atrial pressure was 1.5 m m H g higher in the normal dogs (P< 0.005). Apart from a slight increase in mean right atrial pressure and a decrease in RV minimal diastolic pressure with initial opening of the AV fistula, right heart pressures changed little in either group during aortic pressure reduction (Table 2 ) . T h e cardiac index in normal dogs increased progressively with aortic pressure reduction. I n hypertensive dogs, the cardiac index rose at MAP100 and
R V coronary blood Jow in hypertension
5 11
MAP90 but then tended to fall at MAP80 (Table 2).
Right uentracular blood pow and coronary vascular resistance
Blood flow to the RV free wall in hypertensive dogs (79 ml min 100 g-') was not different from that measured in normal animals (77 ml min 100 g-'). RV coronary blood flow in normotensive dogs at the three reduced pressure levels (73 to 76 ml min 100 g-') was similar to that observed in the control state (Fig. 1a). By contrast, while RV blood flow in hypertensive dogs did not change significantly from control a t either MAP100 (76 ml min 100 g-') or MAP90 (75 ml rnin 100 g-'), it decreased by 18% to 63 ml min 100 g-' at MAP80 ( P < 0.005) (Fig. 1a). This decrease in RV blood flow at MAP80 in hypertensive dogs differed significantly (P < 0.01) from the response observed in normal dogs. Coronary vascular resistance in hypertensive dogs (257 kPa litre-' rnin g-') was slightly but not significantly higher than in the normal animals (234 kPa litre-' min 100 g-') under control conditions. RV coronary vascular resistance decreased progressively in both groups as aortic pressure was lowered to MAP100 and MAP90. However, while RV coronary vascular resistance decreased by about 9% between MAP90 and MAP80 in normal dogs ( P < 0.05), there was no further reduction in the hypertensive animals (Fig. 1 b & c).
Regional right ventricular blood ftow Over the pressures examined, RV endocardial flow exceeded epicardial flow in hypertensive dogs ( P < 0.025). Moreover, the flow reduction at MAP80 was slightly more pronounced in inner muscle layers so that the endocardial-toepicardial flow ratio fell from a range of 1.12-1.16 between control and MAP90 to 1.07 at MAP80 (P< 0.025). In normal dogs, endocardial flow was not significantly different from epicardial flow at any pressure (Fig. 2a) and the endocardial-to-epicardial flow ratio showed only minor fluctuations between 1.05 and 1.08 (Fig. 2 b). In hypertensive dogs, blood flow to full-
I
I
1
I
1
I
80
100
120
140
Mean Aortic Pressure Fig. 1. Changes in average RV blood flow (a), coronary vascular resistance (b) and coronary vascular resistance as a percentage of control (c) in normal and
hypertensive dogs during aortic pressure reduction. N, normotension; HT, hypertension. Bars denote 1 SE of the difference between any 2 means within animals from 2 way ANOVA. *** P < 0.005 compared to flows at higher pressures, 2 way ANOVA. * P < 0.05 for decrease from next highest pressure, 1 tailed paired t test.
thickness segments around the RV free wall under control conditions was highest in the anterior, inferior and posterior margins and lowest in the centre and the adjacent part of the upper margin (Fig. 3). Blood flow did not change 18-2
J.J. Smolich et al.
512
Normotension
(a)
3 ii0
80
-
0 ia
Hypertension
1
u?
0-' 70-
0 '.E m E
EPI
>z OC
60-
0 .c
1.15
(b)
.-
P
e 1.05 U
C
w
I
I
I
I
I
I
I
80
100
120
80
100
120
140
Mean Aortic Pressure Fig. 2. Transmural RV blood flow (a) and endocardial-to-epicardial flow ratio (b) in normal and hypertensive dogs. ENDO, subendocardium; EPI, subepicardium. Bars denote 1 SE of the difference between any 2 means within animals from 2 and 3 way ANOVA. * P < 0.025 compared to flow ratios at higher pressures, 2 way ANOVA.
appreciably in any segment of the RV free wall at either MAP100 or MAP90 in the hypertensive dogs, but fell significantly in all regions at MAP80 (Fig. 3). The distribution of blood flow around the RV free wall in normotensive animals in the control state was similar to that seen in the hypertensive dogs. However, in contrast to the hypertensive animals, blood flow did not change significantly in any segment during aortic pressure reduction in normal dogs (Fig. 3).
Blood Jow to the right side o f the septum Under control conditions in hypertensive animals, blood flow to the right quarter of the septum (117 ml min 100 g-') was greater than that to the endocardia1 half of the RV free wall (85 ml min 100 g-') ( P < 0.005). With aortic pressure reduction in hypertensive dogs, blood flow to both the right quarter of the septum and the RV endocardium was maintained down to MAP90 but fell significantly at MAP80 ( P < 0.005) (Fig. 4a and b). In normotensive dogs,
blood flow to the right side of the septum was also greater than that to the RV subendocardium under control conditions ( P < O.OOS), and both remained constant during lowering of aortic pressure (Fig. 4a and b).
Coronary Artery Morphometry
The internal diameter of the left anterior descending coronary artery segments ranged from 0.42-2.27 mm in hypertensive dogs and from 0.62-2.52mm in normal animals. The cross-sectional area of the media ranged from 0.15-0.87 mm2 and 0.06-0.53 mm2 in hypertensive and normal dogs respectively. The relationship between the cross-sectional area of the media (Y)and the length of the internal elastic lamina (X)was linear in both groups of dogs ( r > 0.99). The slope of this relationship in hypertensive dogs ( Y = - 0.13 0.15X) was steeper than in the normal animals ( Y = -0.01 0.06X) ( P < 0.001), consistent with a greater wall-to-lumen ratio.
+
+
R V coronary blood $ow in hypertension
y PI 76
75
N
HT
N
HT
63 63 65 62 (5)
63 62 62 53x (5)
74 71 71 72 (4)
77 70 69 59ar (5)
71 71 73 71 (6)
76 73 72 61** (5)
77 75 76 78 (7)
(6)
83 88 77 77 (8)
81 78 78 65** (6)
77 75 72
87 81 85
513
84 82 81 66***
Fig. 3. Regional blood flow to full-thickness segments of the RV free wall at the control (C) and reduced pressure levels PI, P, and P, (pressure levels and abbreviations as in Table 2). N, normotension ; HT, hypertension. For each region, values are expressed as mean (1 SE of the difference between any 2 means within animals from 2 way ANOVA). * P < 0.025, **P< 0.005 compared to flows at higher pressures, 2 way ANOVA.
DISCUSSION In this study, we have compared RV coronary blood flow responses during stepwise reduction of aortic pressure in normal dogs and those with LV hypertrophy secondary to chronic renal hypertension. I n normal dogs, RV coronary vascular resistance decreased progressively and RV blood flow was well-maintained as mean aortic pressure was lowered from 128 to 78 mmHg, implying that autoregulation of RV myocardial perfusion was very effective over this pressure range (Smolich et al. 1988a). In hypertensive animals, RV blood flow was initially preserved during reduction of mean aortic pressure from 146 to 91 mmHg. However, as mean aortic pressure was lowered further to 77 mmHg, RV blood flow declined substantially while RV coronary vascular resistance did not decrease significantly. The latter findings suggest that RV blood flow was also autoregulated over a broad pressure range in the hypertensive dogs, but that autoregulation was impaired at the
lowest pressure examined. This conclusion is also consonant with the changes in cardiac index observed in hypertensive dogs in our study (Table 2), because cardiac output initially increases with aortic pressure reduction but then falls below the lower limit of coronary autoregulation (Isoyama et al. 1981). Taken together, our observations are thus consistent with a 'resetting' of the lower end of the RV autoregulatory plateau to a higher pressure level in the hypertensive dogs. Resetting of the lower limit of the autoregulatory plateau to a higher pressure level in chronic hypertension has been reported in the cerebral circulation (Jones et af. 1976, Strandgaard 1976) and, more recently, in the LV coronary circulation (Harrison et al. 1988). Our findings in the right ventricle not only complement the results of the latter study, but also provide an insight into the relative contribution of hypertension-induced vascular changes and cardiac muscle hypertrophy to this alteration of the RV pressure-flow relation. Two observations indicate that the RV muscu-
514
3.3.Smolich 140
4
et al. Normotension
Hypertension
(a)
I1 =
60
./-* I
3
110
R
V
~
~
1
I
(b)
-
0 L
2
c C
90
0
0
8
70 80
100
120
80
100
120
140
Mean Aortic Pressure
Fig. 4. Changes in blood flow (a) and blood flow as a percentage of the control flow (b) in the right quarter of the interventricular septum (RV,) and the inner half of the RV free wall (RV,,) during arterial pressure reduction in normal and hypertensive dogs. Error bars denote 1 SE of the difference between any 2 means within animals from 3 way ANOVA.
lature was relatively unaffected by chronic renovascular hypertension in our study. Firstly, at the same mean left atrial pressure, RV pressures were not elevated compared to those in control animals. Secondly, the postmortem RVto-body weight ratio in hypertensive dogs was similar to that of the control dogs. On the other hand, our morphological observations pointed to appreciable medial hypertrophy in arterial segments, a finding which is in accord with other studies of the vasculature in chronic hypertension (Folkow et al. 1970, Noresson et al. 1977, Greditzer & Fischer 1978, Lee et al. 1983, Owens et al. 1988). Our results, therefore, imply that the resetting of the lower limit of RV autoregulation to a higher pressure level in our study was related predominantly to the effects of chronic hypertension on coronary vascular design. Furthermore, as the RV myocardium of the dog receives a dual blood supply from both the left and right coronary arteries (Donald & Essex 1954, Miller et al. 1964, Murray & Vatner 1980), the similarity of regional blood flow patterns observed in the RV free wall (Fig. 3) and the right side of the septum (Fig. 4) suggests that
this resetting of the RV pressure-flow relation in hypertensive dogs was independent of the anatomical source of the RV arterial supply. Two methodological aspects of our study warrant comment. Firstly, in conjunction with the radioactive microsphere technique, aortic pressure reduction was particularly suitable for evaluating the RV myocardial pressure-flow relation in the dog. In this species the right coronary artery contributes approximately 60 yo of the blood flow to the RV free wall, while branches of the left coronary artery supply the remainder of the RV free wall (Donald & Essex 1954, Murray & Vatner 1980), as well as all of the septum (Miller et af. 1964). Lowering aortic pressure thus decreased left and right coronary arterial pressures simultaneously and thereby produced a uniform reduction in RV perfusion pressure. Our approach thus differed from other studies (Cross 1962, Urabe et al. 1985, Yonekura et al. 1987) which have examined the right coronary artery pressure-flow relation. These studies have found that right coronary artery blood flow decreases linearly with decreases in right coronary artery pressure, implying that RV
RV coronary bloodflow in hypertension autoregulatory capacity is limited. However, as complete right coronary artery occlusion has only a partial effect on RV myocardial blood flow in the dog (Murray & Vatner 1980), it appears that changes in right coronary artery flow may not necessarily reflect alterations in RV myocardial perfusion in this species. This apparent discrepancy is probably related to the presence of 'overlap flow' (Marcus 1983) arising from extensive Rv interdigitation of left and right coronary arterial branches. Secondly, experiments in normal and hypertensive dogs were performed under highly controlled conditions to minimize changes in RV metabolism. Thus, aortic pressure was lowered without changes in heart rate (Nakano & D e Schryver 1964, Fujisawa et al. 1984), cardiac contractility (Fujisawa et al. 1984) and filling pressures (Nakano & De Schryver 1964) which normally accompany opening of an AV fistula. Aortic pressure reduction was associated with rises in cardiac output in our study. However, on the basis of data available for the left ventricle (Sarnoff et al. 1958, Taylor et al. 1967, Suga et al. 1982), the associated increase in RV muscle shortening would have only increased RV oxygen requirements and blood flow by 2 4 % in the hypertensive dogs and 3-6 yo in the normal dogs between the control and highest level of cardiac output. I n summary, RV coronary blood flow was measured with radioactive microspheres during moderate aortic pressure reduction in anaesthetized, autonomically blocked normal and renal hypertensive dogs. I n normal dogs, total and regional RV blood flow remained constant at all pressures examined. By contrast, total and regional RV blood flow was initially maintained in hypertensive dogs during aortic pressure reduction, but then fell substantially. We conclude firstly, that autoregulation of RV blood flow was preserved in dogs with chronic hypertension but that, compared to normal dogs, the lower limit of autoregulation was reset to a higher pressure level ; secondly, that this impairment of pressure-dependent vasodilation at moderately low perfusion pressures was probably related to hypertension-induced structural changes in the coronary vasculature. We are grateful to Dr Archer Broughton for his constructive input into this study. We also thank Anna-Lisa Leb, Julie Safstrom, Kathryn Pempel and
515
Ruta Dumpys for their technical assistance, and Dr Frances Cribbin and Tessa Morton for technical editing of the manuscript. J.J. S . was supported by the National Heart Foundation of Australia. P. L. W. was the recipient of a British Medical Research Council Travelling Fellowship. P. F. was a Postdoctoral Research Fellow of the Swedish Medical Research Council.
REFERENCES ANDERSON,W.P., KORNER, P.I. & JOHNSTON, C.I. 1979. Acute angiotensin 11-mediated restoration of distal renal artery pressure in renal artery stenosis and its relationship to the development of sustained one-kidney hypertension in conscious dogs. Hypertension 1 , 292-298. BROUGHTON, A. & KORNER,P.I. 1983. Basal and maximal inotropic state in renal hypertensive dogs with cardiac hypertrophy. Am 3 Physiol 245, H33-H41. CROSS,C.E. 1962. Right ventricular pressure and coronary flow. Am 3 Physiol202, 12-16. DONALD,D.E. & ESSEX,H.E. 1954. Pressure studies after inactivation of the major portion of the canine right ventricle. Am 3 Physiol 176, 155-161. FOLKOW, B., HALLBACK, M., LUNDGREN, Y. & WEISS, L. 1970. Structurally based increase of flow resistance in spontaneously hypertensive rats. Acta Physiol Scand 79, 373-378. FUJISAWA, A,, SASAYAMA, S., TAKAHASHI, M., NAKAMURA, M., OHYAGI,A,, LEE, J-D., YUI, Y. & KAWAI,C. 1984. Enhancement of left ventricular contracility after opening of an arteriovenous fistula in dogs. Cardiovasc Res 18, 51-59. GREDITZER, H.G. 111 & FISCHER, V.W. 1978. A sequential ultrastructural study of different arteries in the hypertensive rat. Ex$ Mol Pathol29, 12-28. GREENSMITH, J.E. & DULING, B.R. 1984. Morphology of the constricted arteriolar wall : physiological implications. Am 3 Physiol 247, H687-H698. D.G., FLORENTINE, M.S., BROOKS, L.A., HARRISON, M.L. 1988. The effect of COOPER, S.M. & MARCUS, hypertension and left ventricular hypertrophy on the lower range of coronary autoregulation. Circulation 77, 1108-1115. HEYMANN, M.A., PAYNE,B.D., HOFFMAN, J.I.E. & RUDOLPH,A.M. 1977. Blood flow measurements with radionuclide-labeled particles. Prog Cardiovasc Dis 20, 55-79. S., MARUYAMA, Y., KOIWA, Y., ISHIDE,N., ISOYAMA, KITAOKA, S., TAMAKI, K., SATO,S., SHIMIZU, Y., INO-OKA,E. & TAKISHIMA, T . 1981. Experimental study of afterload-reducing therapy. The effects of the reduction of systemic vascular resistance on cardiac output, aortic pressure and coronary circulation in isolated, ejecting canine hearts. Cwculataon 64, 490-499.
516
3.3. Smolich
et al.
J.V., FITCH, W., MACKENZIE, E.T., STRAND- pressure reduction in the autonomically blocked dog : evidence for right ventricular autoregulation. S. & HARPER,A.M. 1976. Lower limit of Cardiovasc Re5 22, 17-24. cerebral blood flow autoregulation in experimental J.J., WEISSBERG, P.L., BROUGHTON, A. & renovascular hypertension in the baboon. Circ Res SMOLICH, KORNER,P.I. 1988b. Aortic pressure reduction 39, 555-557. redistributes transmural blood flow in dog left KORNER, P.I. & HILDER,R.G. 1974. Measurement of ventricle. A m 3 Physiol 254, H361-H368. cardiac output and regional blood flow by thermoG.W. & COCHRAN, W.G. 1980. Statistical dilution. Clin Exp Pharmacol Physiol Suppl 1, SNEDECOR, methods. 7th ed. Iowa State University Press, Iowa. 47-66. S. 1976. Autoregulation of cerebral STRANDGAARD, LEE,R.M.K.W., FORREST, J.B., GARFIELD, R.E. & blood flow in hypertensive patients. The modifying DANIEL,E.E. 1983. Comparison of blood vessel1 influence of prolonged antihypertensive treatment wall dimensions in normotensive and hypertensive on the tolerance to acute, drug-induced hyporats by histometric and morphometric methods. tension. Circulation 53, 720-727. Blood Vessels 20, 245-254. SUGA,H., HISANO, R., HIRATA, S., HAYASHI, T. & MARCUS, M.L. 1983 The coronary circulation in health NINOMIYA, I. 1982. Mechanism of higher oxygen and disease. McGraw-Hill, New York. consumption rate : pressure-loaded vs. volumeMILLER,M.E., CHRISTENSEN, G.C. & EVANS,H.E. loaded heart. A m 3 Physiol 242, H942-H948. 1964. Anatomy of the dog, pp 283-285. WB TAYLOR, R.R., CINGOLANI, H.E., GRAHAM, T.P. & Saunders, Philadelphia. CLANCY, R.L. 1967. Myocardial oxygen consumpMURRAY,P.A. & VATNER,S.F. 1980. Fractional tion, left ventricular fibre shortening and wall contributions of the right and left coronary arteries tension. Cardiovasc Res 1, 219-228. to perfusion of normal and hypertrophied right TOMANEK, R.J., WANGLER, R.D. & BAUER, C.A. 1985. Prevention of coronary vasodilator reserve decventricles of conscious dogs. Circ Res 47, 190-200. rement in spontaneously hypertensive rats. HyNAKANO, J. & DE SCHRYVER, C. 1964. Effects of pertension 7, 533-540. arteriovenous fistula on systemic and pulmonary URABE,Y., TOMOIKE, H., OHZONO,K., KOYANAGI, S. circulations. A m 3 Physiol 207, 1319-1324. & NAKAMURA, M. 1985. Role of afterload in NORESSON, E., HALLBACK, M. & HJALMARSSON, A. determining regional right ventricular performance 1977. Structural ‘resetting’ of the coronary vascular during coronary underperfusion in dogs. Circ Res, bed in spontaneously hypertensive rats. Acta Physiol 57, 9&104. Scand 101, 363-365. R. D., PETERS,K. G., MARCUS, M. L. & OWENS,G.K., SCHWARTZ, S.M. & MCCANNA,M. WANGLER, TOMANEK, R. J. 1982. Effects of duration and 1988. Evaluation of medial hypertrophy in reseverity of arterial hypertension and cardiac hypersistance vessels of spontaneously hypertensive rats. trophy on coronary vasodilator reserve. Circ Res Hypertension 11, 198-207. 51, 10-18. SARNOFF, S.J., BRAUNWALD, E., WELCH,G.H. JR., WICKER, P. & TARAZI, R. C. 1985. Right ventricular W.N. & MACRUZ, R. 1958. CASE,R.B., STAINSBY, coronary flow in arterial hypertension. A m Heart 3 Hemodynamic determinants of oxygen consump110, 845-850. tion of the heart with special reference to the YONEKURA, S., WATANABE, N., CAFFREY, J.L., GAUGL, tension-time index. A m 3 Physiol 192, 148-156. J.F. & DOWNEY, H.F. 1987. Mechanism of attenuSMOLICH, J.J., WEISSBERG, P.L., BROUGHTON, A. & ated pressure-flow autoregulation in right coronary KORNER, P.I. 1988a. Comparison of left and right circulation of dogs. Circ Res 60, 133-141. ventricular blood flow responses during arterial
JONES,
GAARD,
ADONIS
0001667729100075G
Right ventricular coronary blood flow patterns during aortic pressure reduction in renal hypertensive dogs J. J. S M O L I C H , P. L. W E I S S B E R G , P. F R I B E R G and P. I. KORNER Baker Medical Research Institute, Melbourne, Victoria, Australia
SMOLICH, J. J., WEISSBERG, P. L., FRIEERG, P. & KORNER, P. I. 1991. Right ventricular coronary blood flow patterns during aortic pressure reduction in renal hypertensive dogs. Acta Ph,ysiol Scund 141, 507-516. Received 17 July 1990, accepted 22 October 1990. ISSN 0001-6772. Baker Medical Research Institute, Melbourne, Victoria, Australia. We measured right ventricular coronary blood flow with radioactive microspheres during graded aortic pressure reduction in 13 normal dogs and in 13 renal hypertensive dogs with left ventricular hypertrophy. Under anaesthesia and controlled loading conditions, mean aortic pressure was lowered from control (128 mmHg in normal and 146 mmHg in hypertensive dogs) to approximately 100, 90 and 80 mmHg. In normal dogs, right ventricular blood flow was not affected by this pressure reduction, consistent with effective right ventricular autoregulation. In hypertensive dogs, however, right ventricular blood flow was maintained between a mean aortic pressure of 146 and 90 mmHg (range 75-79 ml min-' 100 g-') but fell by 18% to 63 ml min 100 g-' at a mean aortic pressure of 80 mmHg ( P < 0.005). We conclude that autoregulation of right ventricular blood flow was preserved in chronic hypertension but that, compared to normal dogs, the lower limit of autoregulation was reset to a higher pressure level. Moreover, the similarity of right ventricular-to-body weight ratios in the two groups implied that this change was a consequence of hypertension-induced structural changes in the coronary vasculature. Key words ; autoregulation ; coronary circulation ; coronary vascular resistance ; hypertension ; radioactive microspheres.
T h e left ventricle (LV) exposed to chronic systemic hypertension commonly exhibits coronary blood flow abnormalities (Marcus 1983). However, because of the coexistence of hypertension-induced vascular changes and muscle hypertrophy in the left ventricle in this condition, it may be difficult to separate the relative contribution of these two factors to any alterations in LV blood flow patterns. T h i s is less of a problem in the right ventricle (RV) because, while the coronary vasculature of this chamber is exposed to the same elevated arterial pressures as the left ventricle, RV muscle hypertrophy is Correspondence : Dr J. J. Smolich, Centre for Early Human Development, Monash Medical Centre, Clayton Road, Clayton, Victoria, Australia, 3 168.
either absent (Tomanek et al. 1985) or of only minor degree (Wangler et al. 1982, Wicker & Tarazi 1985). Previous coronary flow studies of the right ventricle in chronic hypertension (Wangler e t al. 1982, Tomanek et al. 1985, Wicker & Tarazi 1985), have demonstrated that RV blood flow is preserved under control conditions, even though RV coronary vascular resistance is increased. Little is known, however, about the effects of systemic hypertension on RV blood flow patterns during changes in RV perfusion pressure. I n an earlier study in normal dogs (Smolich et al. 1988a), we observed that RV myocardial perfusion, measured with radioactive microspheres, remained essentially constant during reductions in RV perfusion pressure produced
507
508
3.3. Smolich et al.
b y stepwise lowering of aortic pressure. The aim of the present study was to determine the effect of renovascular hypertension on RV blood flow responses during aortic pressure reduction.
MATERIALS AND METHODS Animals. T h e full study protocol was performed in 26 mongrel dogs. T h e experimental group comprised 13 dogs with renovascular hypertension. T h e 13 dogs in the control group were selected on the basis of a similar range of body weights and were studied concurrently with the hypertensive group. All animals received several weeks of health conditioning prior to any surgical procedure. All experimental procedures were in accord with guidelines set by the National Health and Medical Research Council of Australia. Induction of hypertension. One-kidney renovascular hypertension was induced as previously described (Anderson et al. 1979, Broughton & Korner 1983). Briefly, animals were anaesthetized with i.v. sodium thiopentone, intubated and ventilated with a mixture of oxygen, nitrous oxide and halothane. Indwelling polyvinyl catheters were inserted into the abdominal aorta through a left flank incision. An inflatable Silastic cuff (Hazen Everett, New Jersey, USA) and a continuous wave Doppler flow probe were placed around the left renal artery. The right kidney was then removed through a right flank incision. Postoperatively, animals were given oral antibiotics for 7 days and analgesia as required. Recovery from surgery was complete by 1&14 days. Thereafter, aortic blood pressure was measured with the animals fully conscious and unrestrained but lying quietly on a padded table. A gradual rise in arterial blood pressure over several weeks was produced by progressive inflation of the renal artery cuff. The elevated blood pressure was maintained for at least 6 week prior to the acute experiment. Acute experiment. Normal and hypertensive dogs were anaesthetized with i.v. sodium pentobarbitone (20-30 mg kg bolus followed by a constant infusion of 6 mg kg h). The lungs were ventilated at an endexpiratory pressure of 4 cm H,O (0.4kPa) with oxygen-enriched air. Arterial blood gases were checked frequently (Radiometer ABL1, Copenhagen, Denmark) and ventilation was adjusted to maintain P,,, between 35 and 40 mmHg (4.7-5.3 kPa) and arteriaf p H between 7.3 and 7.4. Base deficits were corrected with i.v. sodium bicarbonate as required. Body temperature was maintained between 36 "C and 38 "C with a heated table and water blanket. T h e surgical preparation of the animals was as previously described (Smolich et al. 1988a, 1988b). The chest was opened in the 5th left interspace. Fluid-filled pressure catheters were inserted into the left atrium, and into the ascending aorta via the
carotid artery. A large cannula was introduced into the left atrium through the appendage and connected to a reservoir filled with a mixture of room-temperature heparinized blood and low molecular weight dextran (10% Rheomacrodex in normal saline). A highfidelity micromanometer (Konigsberg P4.5) was calibrated in a 37 "C water bath against a mercury manometer and then inserted into the left ventricle through the apical dimple in all animals as part of another protocol. A second micromanometer (Konigsberg P4.5) was introduced into the RV cavity through its free wall to measure RV pressure in a limited number of control and hypertensive animals. In these animals a catheter was also inserted into the right atrial cavity and/or the pulmonary trunk to provide a pressure reference for the RV micromanometer. The edges of the pericardial incisions were loosely approximated with interrupted sutures. Lastly, a plastic Ypiece was inserted into the thoracic aorta during several minutes occlusion between cross-clamps. An arteriovenous (AV) fistula was created by connecting the protruding arm of the Y-piece to a jugular venous cannula with wide-bore tubing. Flow through the fistula was regulated with a roller pump. After the completion of surgery, the chest wound was left open but covered with towels to minimize heat loss. Experimental protocol. Autonomic ganglionic transmission was blocked with mecamylamine (2 mg kg-' i.v.) and the cervical vagosympathetic trunks were cut to eliminate reflex changes in cardiac contractility arising from reductions in aortic pressure (Broughton & Korner 1983, Smolich et al. 19886). The heart was paced at 150 beats min-' with a stimulator (Grass SD9) and bipolar left atrial electrode. Arterial and atrial pressures were measured with Statham P23D strain gauge transducers using the mid-thoracic vertebral spines as the zero reference level. Openchest mean left atrial pressure was held at 15 mmHg with the left atrial reservoir. Ventricular, arterial and atrial pressures were displayed on a Devices MI6 paper recorder. I n both the normal and hypertensive dogs, mean aortic pressure (MAP) was reduced in steps with the AV fistula from the control value to approximately 100 (MAPlOO), 90 (MAP90) and 80 mmHg (MAP80). Steady state conditions were maintained for several min at each reduced pressure level. Haemodynamic variables were recorded and myocardial blood flow then measured at the control and reduced pressure levels with 15 ,um radioactive microspheres (Heymann et al. 1977). Isotopes were randomly selected from six available labels (141Ce %r l13Sn "Sr 95Nb, 46Sc; New England Nuclear). Approximately 1 million microspheres were mixed in 2 ml of physiological saline for 10 min by ultrasonification and manual agitation and then injected into the left atrial cavity over 30-40 s with 15 ml of saline. A reference sample was withdrawn at 12 ml/min from an aortic catheter
RV coronary bloodjow in hypertension with a mechanical pump (Harvard 940A, Massachusetts, USA), starting 5-10 s before injection and ending 75-90 s after injection. Cardiac output was measured with the thermodilution technique (Korner & Hilder 1974) in seven normal and seven hypertensive dogs during the pressure reduction. Room-temperature 5 yo dextrose was injected into the right atrium and the resultant time-temperature curves were recorded with a thermistor probe situated in the descending aorta or a major pulmonary artery. Cardiac output was determined in duplicate and indexed for body wt. RV and mean right atrial pressures were also measured in a separate group of three normal animals which underwent aortic pressure reduction according to the above protocol, but which did not receive the full complement of microsphere injections. Fixation of hearts. At the end of the experiment, two normal and two hypertrophied hearts were arrested in diastole with KCI, cleared of blood and fixed at a perfusion pressure of 100 mmHg with 2% paraformaldehyde and 2% glutaraldehyde in 0.1 M phosphate buffer. Segments of the left anterior descending coronary artery were excised from proximal, mid and distal sites. These were fixed for a further 2 h in 1% osmium tetroxide in 0.1 M phosphate buffer, dehydrated in acetone and embedded in eponaraldite. Semi-thin sections were cut with glass knives and stained with 1% methylene blue. The sections were projected onto a computer-linked digitizing tablet and the cross-sectional area of the media and the length of the internal elastic lamina measured. As neither are appreciably influenced by vessel contraction (Lee et al. 1983, Greensmith & Duling 1984), linear regression of medial cross-sectional area on the length of the internal elastic lamina provided a measure of the wall-to-lumen ratio which was not affected by any fixation-related differences in vessel contraction between normal and hypertensive animals (Owens et al. 1988). The other dogs were sacrificed with an overdose of sodium pentobarbitone and potassium chloride at the end of the experiments and the hearts immersed in fixative for 7-10 days. Myocardial blood jlow measurements. The large coronary vessels, epicardial fat and valves were removed. The RV free wall was separated from the septum and dissected into 9 endocardial and 9 epicardial samples (Smolich et al. 1988a). In the normal dogs, the former constituted 40.1 f0.9% and the latter 59.9f0.9y0 of the total RV weight. The corresponding contributions of these two layers in the hypertensive dogs were 39.6 f0.7% and 60.4 0.7% respectively. The right quarter of the interventricular septum, which is anatomically continuous with the RV subendocardium, was also removed and divided into 6 segments. 18
509
Isotope radioactivity in the reference (r) and tissue (t) samples was determined with a Packard 5130 through-hole gamma counter. The radioactivity of each isotope peak in the tissue samples was obtained by stripping away the contribution of background and the Compton areas of any higher energy isotopes with a computer program. Blood flow to the tissue samples was calculated from the relation: Q, = (Qx R,)/R,, where Q is flow (ml min-') and R is radioactivity (counts min-l). Total flow to the RV free wall and the right side of the interventricular septum was obtained by summing the individual sample flows. Flows to the various full-thickness regions, and the endocardial and epicardial halves of the RV free wall were calculated by appropriate combination of the same sample flows. All flows were normalized to 100 g of wet tissue. Coronary vascular resistance. Relative coronary vascular resistance of the RV free wall was calculated where QM,,is average RV from (PA,, - PA,,)/&,,,, blood flow (ml min 100 g-'), PART is mean arterial pressure (kPa) and PA,, is mean right arterial pressure (kPa). Mean right atrial pressure was assumed to be 11 mmHg (1.5 kPa) in control dogs and 9 mmHg (1.2 kPa) in the hypertensive animals (Table 2). Statistical analysis. Results were evaluated with standard statistical tests (Snedecor & Cochran 1980). Changes in haemodynamic variables and RV blood flow in normal and hypertensive dogs at the various pressure levels were analyzed with two and three way analysis of variance (ANOVA). The between pressures and, where appropriate, the between regions sums of squares was orthogonally partitioned into individual degrees of freedom. Differences between normal (N) and hypertensive (HT) dogs were evaluated with t = A(HT - N)/(SEi SE;,) 0.5 where A(HT - N) was the difference between the response of normal and hypertensive dogs and SEN and SE,, were the appropriate standard errors from analysis of variance (Broughton & Korner 1983). The significance of decreases in RV coronary vascular resistance during aortic pressure reduction was assessed with a one tailed paired t-test. Differences in the relationship between medial cross-sectional area and the length of the internal elastic lamina in coronary arteries of normal and hypertensive dogs were analyzed with analysis of covariance, using the pooled regression coefficient of each group. A value of P < 0.05 was considered significant. Unless specified otherwise, results are expressed as mean SE.
+
RESULTS Mean arterial pressure rose from an average baseline value of 103f2 m m H g to 131f4 mm1 ig following renal artery constriction in the experimental group of dogs. T h e elevated blood ACT 141
510
3.3.Smolich et al.
Table 1. Body and ventricular weights in the normal and hypertensive dogs
BW, kg LVW/BW, g/kg RVW/BW, g/kg
Normal ( n = 13)
Hypertension ( n = 13)
27.4f 1.7 4.16k0.16 1.30f0.05
27.3 f 1.2 5.13 f0.23 1.36k0.04
P NS
< 0.005 NS
BW, body weight at time of acute experiment; LVW, left ventricular weight; RVW, right ventricular weight. P, normal us. hypertension, unpaired two tailed t-test ; NS, not significant.
Table 2. Haemodynamic parameters in the control state (C) and three reduced pressure levels (PI, P,, P3) in normal and hypertensive dogs C MAP, mmHg Normal ( n = 13) Hypertension ( n = 13) RVSP, mmHg Normal (n = 6) Hypertension (n = 7) RVDP, mmHg Normal ( n = 6) Hypertension ( n = 7) MRAP, mmHg Normal ( n = 6) Hypertension ( n = 5) CI, ml min kg-' Normal (n = 7) Hypertension ( n = 7)
SE
p3
Pl
____
128 146"*
104 107
91 91
78 77
3 4
31 31
32 30
31 30
31 31
1 1
5
6 6
5 6
5 7
1 1
10.8 9.0
10.9 9.0
11.0 9.3
0.4 0.3
8XXX 10.2" 9.3 149 157
175 179
195 199
202 187
11 9
MAP, mean aortic pressure ; RVSP, right ventricular systolic pressure; RVDP, right ventricular minimal diastolic pressure; MRAP, mean right atrial pressure; CI, cardiac index; SE, SE of the difference between any two means within animals from two-way ANOVA. ** P < 0.025, normal us. hypertension, unpaired t-test. * P < 0.05, *"" P < 0.005, compared to reduced pressure levels, two-way ANOVA.
pressure was maintained for an average of 10 week (range 6-17 week). At postmortem, the LV-to-body weight ratio of the hypertensive dogs was 23 yogreater than in normal dogs (P< O.OOS), but the RV-to-body weight ratios were closely similar in the two groups (Table 1).
Haemodynamic indices During the initial control period under anaesthesia, mean aortic pressure in the hypertensive animals was significantly higher than in the normal dogs, but this variable was matched in the two groups at lower pressures (Table 2). Phasic RV pressures were similar in hypertensive
and normal dogs at all pressures. With an identical mean left atrial pressure, mean right atrial pressure in normal and hypertensive dogs did not differ significantly under control conditions. However, when the grouped values from all aortic pressure levels were compared, mean right atrial pressure was 1.5 m m H g higher in the normal dogs (P< 0.005). Apart from a slight increase in mean right atrial pressure and a decrease in RV minimal diastolic pressure with initial opening of the AV fistula, right heart pressures changed little in either group during aortic pressure reduction (Table 2 ) . T h e cardiac index in normal dogs increased progressively with aortic pressure reduction. I n hypertensive dogs, the cardiac index rose at MAP100 and
R V coronary blood Jow in hypertension
5 11
MAP90 but then tended to fall at MAP80 (Table 2).
Right uentracular blood pow and coronary vascular resistance
Blood flow to the RV free wall in hypertensive dogs (79 ml min 100 g-') was not different from that measured in normal animals (77 ml min 100 g-'). RV coronary blood flow in normotensive dogs at the three reduced pressure levels (73 to 76 ml min 100 g-') was similar to that observed in the control state (Fig. 1a). By contrast, while RV blood flow in hypertensive dogs did not change significantly from control a t either MAP100 (76 ml min 100 g-') or MAP90 (75 ml rnin 100 g-'), it decreased by 18% to 63 ml min 100 g-' at MAP80 ( P < 0.005) (Fig. 1a). This decrease in RV blood flow at MAP80 in hypertensive dogs differed significantly (P < 0.01) from the response observed in normal dogs. Coronary vascular resistance in hypertensive dogs (257 kPa litre-' rnin g-') was slightly but not significantly higher than in the normal animals (234 kPa litre-' min 100 g-') under control conditions. RV coronary vascular resistance decreased progressively in both groups as aortic pressure was lowered to MAP100 and MAP90. However, while RV coronary vascular resistance decreased by about 9% between MAP90 and MAP80 in normal dogs ( P < 0.05), there was no further reduction in the hypertensive animals (Fig. 1 b & c).
Regional right ventricular blood ftow Over the pressures examined, RV endocardial flow exceeded epicardial flow in hypertensive dogs ( P < 0.025). Moreover, the flow reduction at MAP80 was slightly more pronounced in inner muscle layers so that the endocardial-toepicardial flow ratio fell from a range of 1.12-1.16 between control and MAP90 to 1.07 at MAP80 (P< 0.025). In normal dogs, endocardial flow was not significantly different from epicardial flow at any pressure (Fig. 2a) and the endocardial-to-epicardial flow ratio showed only minor fluctuations between 1.05 and 1.08 (Fig. 2 b). In hypertensive dogs, blood flow to full-
I
I
1
I
1
I
80
100
120
140
Mean Aortic Pressure Fig. 1. Changes in average RV blood flow (a), coronary vascular resistance (b) and coronary vascular resistance as a percentage of control (c) in normal and
hypertensive dogs during aortic pressure reduction. N, normotension; HT, hypertension. Bars denote 1 SE of the difference between any 2 means within animals from 2 way ANOVA. *** P < 0.005 compared to flows at higher pressures, 2 way ANOVA. * P < 0.05 for decrease from next highest pressure, 1 tailed paired t test.
thickness segments around the RV free wall under control conditions was highest in the anterior, inferior and posterior margins and lowest in the centre and the adjacent part of the upper margin (Fig. 3). Blood flow did not change 18-2
J.J. Smolich et al.
512
Normotension
(a)
3 ii0
80
-
0 ia
Hypertension
1
u?
0-' 70-
0 '.E m E
EPI
>z OC
60-
0 .c
1.15
(b)
.-
P
e 1.05 U
C
w
I
I
I
I
I
I
I
80
100
120
80
100
120
140
Mean Aortic Pressure Fig. 2. Transmural RV blood flow (a) and endocardial-to-epicardial flow ratio (b) in normal and hypertensive dogs. ENDO, subendocardium; EPI, subepicardium. Bars denote 1 SE of the difference between any 2 means within animals from 2 and 3 way ANOVA. * P < 0.025 compared to flow ratios at higher pressures, 2 way ANOVA.
appreciably in any segment of the RV free wall at either MAP100 or MAP90 in the hypertensive dogs, but fell significantly in all regions at MAP80 (Fig. 3). The distribution of blood flow around the RV free wall in normotensive animals in the control state was similar to that seen in the hypertensive dogs. However, in contrast to the hypertensive animals, blood flow did not change significantly in any segment during aortic pressure reduction in normal dogs (Fig. 3).
Blood Jow to the right side o f the septum Under control conditions in hypertensive animals, blood flow to the right quarter of the septum (117 ml min 100 g-') was greater than that to the endocardia1 half of the RV free wall (85 ml min 100 g-') ( P < 0.005). With aortic pressure reduction in hypertensive dogs, blood flow to both the right quarter of the septum and the RV endocardium was maintained down to MAP90 but fell significantly at MAP80 ( P < 0.005) (Fig. 4a and b). In normotensive dogs,
blood flow to the right side of the septum was also greater than that to the RV subendocardium under control conditions ( P < O.OOS), and both remained constant during lowering of aortic pressure (Fig. 4a and b).
Coronary Artery Morphometry
The internal diameter of the left anterior descending coronary artery segments ranged from 0.42-2.27 mm in hypertensive dogs and from 0.62-2.52mm in normal animals. The cross-sectional area of the media ranged from 0.15-0.87 mm2 and 0.06-0.53 mm2 in hypertensive and normal dogs respectively. The relationship between the cross-sectional area of the media (Y)and the length of the internal elastic lamina (X)was linear in both groups of dogs ( r > 0.99). The slope of this relationship in hypertensive dogs ( Y = - 0.13 0.15X) was steeper than in the normal animals ( Y = -0.01 0.06X) ( P < 0.001), consistent with a greater wall-to-lumen ratio.
+
+
R V coronary blood $ow in hypertension
y PI 76
75
N
HT
N
HT
63 63 65 62 (5)
63 62 62 53x (5)
74 71 71 72 (4)
77 70 69 59ar (5)
71 71 73 71 (6)
76 73 72 61** (5)
77 75 76 78 (7)
(6)
83 88 77 77 (8)
81 78 78 65** (6)
77 75 72
87 81 85
513
84 82 81 66***
Fig. 3. Regional blood flow to full-thickness segments of the RV free wall at the control (C) and reduced pressure levels PI, P, and P, (pressure levels and abbreviations as in Table 2). N, normotension ; HT, hypertension. For each region, values are expressed as mean (1 SE of the difference between any 2 means within animals from 2 way ANOVA). * P < 0.025, **P< 0.005 compared to flows at higher pressures, 2 way ANOVA.
DISCUSSION In this study, we have compared RV coronary blood flow responses during stepwise reduction of aortic pressure in normal dogs and those with LV hypertrophy secondary to chronic renal hypertension. I n normal dogs, RV coronary vascular resistance decreased progressively and RV blood flow was well-maintained as mean aortic pressure was lowered from 128 to 78 mmHg, implying that autoregulation of RV myocardial perfusion was very effective over this pressure range (Smolich et al. 1988a). In hypertensive animals, RV blood flow was initially preserved during reduction of mean aortic pressure from 146 to 91 mmHg. However, as mean aortic pressure was lowered further to 77 mmHg, RV blood flow declined substantially while RV coronary vascular resistance did not decrease significantly. The latter findings suggest that RV blood flow was also autoregulated over a broad pressure range in the hypertensive dogs, but that autoregulation was impaired at the
lowest pressure examined. This conclusion is also consonant with the changes in cardiac index observed in hypertensive dogs in our study (Table 2), because cardiac output initially increases with aortic pressure reduction but then falls below the lower limit of coronary autoregulation (Isoyama et al. 1981). Taken together, our observations are thus consistent with a 'resetting' of the lower end of the RV autoregulatory plateau to a higher pressure level in the hypertensive dogs. Resetting of the lower limit of the autoregulatory plateau to a higher pressure level in chronic hypertension has been reported in the cerebral circulation (Jones et af. 1976, Strandgaard 1976) and, more recently, in the LV coronary circulation (Harrison et al. 1988). Our findings in the right ventricle not only complement the results of the latter study, but also provide an insight into the relative contribution of hypertension-induced vascular changes and cardiac muscle hypertrophy to this alteration of the RV pressure-flow relation. Two observations indicate that the RV muscu-
514
3.3.Smolich 140
4
et al. Normotension
Hypertension
(a)
I1 =
60
./-* I
3
110
R
V
~
~
1
I
(b)
-
0 L
2
c C
90
0
0
8
70 80
100
120
80
100
120
140
Mean Aortic Pressure
Fig. 4. Changes in blood flow (a) and blood flow as a percentage of the control flow (b) in the right quarter of the interventricular septum (RV,) and the inner half of the RV free wall (RV,,) during arterial pressure reduction in normal and hypertensive dogs. Error bars denote 1 SE of the difference between any 2 means within animals from 3 way ANOVA.
lature was relatively unaffected by chronic renovascular hypertension in our study. Firstly, at the same mean left atrial pressure, RV pressures were not elevated compared to those in control animals. Secondly, the postmortem RVto-body weight ratio in hypertensive dogs was similar to that of the control dogs. On the other hand, our morphological observations pointed to appreciable medial hypertrophy in arterial segments, a finding which is in accord with other studies of the vasculature in chronic hypertension (Folkow et al. 1970, Noresson et al. 1977, Greditzer & Fischer 1978, Lee et al. 1983, Owens et al. 1988). Our results, therefore, imply that the resetting of the lower limit of RV autoregulation to a higher pressure level in our study was related predominantly to the effects of chronic hypertension on coronary vascular design. Furthermore, as the RV myocardium of the dog receives a dual blood supply from both the left and right coronary arteries (Donald & Essex 1954, Miller et al. 1964, Murray & Vatner 1980), the similarity of regional blood flow patterns observed in the RV free wall (Fig. 3) and the right side of the septum (Fig. 4) suggests that
this resetting of the RV pressure-flow relation in hypertensive dogs was independent of the anatomical source of the RV arterial supply. Two methodological aspects of our study warrant comment. Firstly, in conjunction with the radioactive microsphere technique, aortic pressure reduction was particularly suitable for evaluating the RV myocardial pressure-flow relation in the dog. In this species the right coronary artery contributes approximately 60 yo of the blood flow to the RV free wall, while branches of the left coronary artery supply the remainder of the RV free wall (Donald & Essex 1954, Murray & Vatner 1980), as well as all of the septum (Miller et af. 1964). Lowering aortic pressure thus decreased left and right coronary arterial pressures simultaneously and thereby produced a uniform reduction in RV perfusion pressure. Our approach thus differed from other studies (Cross 1962, Urabe et al. 1985, Yonekura et al. 1987) which have examined the right coronary artery pressure-flow relation. These studies have found that right coronary artery blood flow decreases linearly with decreases in right coronary artery pressure, implying that RV
RV coronary bloodflow in hypertension autoregulatory capacity is limited. However, as complete right coronary artery occlusion has only a partial effect on RV myocardial blood flow in the dog (Murray & Vatner 1980), it appears that changes in right coronary artery flow may not necessarily reflect alterations in RV myocardial perfusion in this species. This apparent discrepancy is probably related to the presence of 'overlap flow' (Marcus 1983) arising from extensive Rv interdigitation of left and right coronary arterial branches. Secondly, experiments in normal and hypertensive dogs were performed under highly controlled conditions to minimize changes in RV metabolism. Thus, aortic pressure was lowered without changes in heart rate (Nakano & D e Schryver 1964, Fujisawa et al. 1984), cardiac contractility (Fujisawa et al. 1984) and filling pressures (Nakano & De Schryver 1964) which normally accompany opening of an AV fistula. Aortic pressure reduction was associated with rises in cardiac output in our study. However, on the basis of data available for the left ventricle (Sarnoff et al. 1958, Taylor et al. 1967, Suga et al. 1982), the associated increase in RV muscle shortening would have only increased RV oxygen requirements and blood flow by 2 4 % in the hypertensive dogs and 3-6 yo in the normal dogs between the control and highest level of cardiac output. I n summary, RV coronary blood flow was measured with radioactive microspheres during moderate aortic pressure reduction in anaesthetized, autonomically blocked normal and renal hypertensive dogs. I n normal dogs, total and regional RV blood flow remained constant at all pressures examined. By contrast, total and regional RV blood flow was initially maintained in hypertensive dogs during aortic pressure reduction, but then fell substantially. We conclude firstly, that autoregulation of RV blood flow was preserved in dogs with chronic hypertension but that, compared to normal dogs, the lower limit of autoregulation was reset to a higher pressure level ; secondly, that this impairment of pressure-dependent vasodilation at moderately low perfusion pressures was probably related to hypertension-induced structural changes in the coronary vasculature. We are grateful to Dr Archer Broughton for his constructive input into this study. We also thank Anna-Lisa Leb, Julie Safstrom, Kathryn Pempel and
515
Ruta Dumpys for their technical assistance, and Dr Frances Cribbin and Tessa Morton for technical editing of the manuscript. J.J. S . was supported by the National Heart Foundation of Australia. P. L. W. was the recipient of a British Medical Research Council Travelling Fellowship. P. F. was a Postdoctoral Research Fellow of the Swedish Medical Research Council.
REFERENCES ANDERSON,W.P., KORNER, P.I. & JOHNSTON, C.I. 1979. Acute angiotensin 11-mediated restoration of distal renal artery pressure in renal artery stenosis and its relationship to the development of sustained one-kidney hypertension in conscious dogs. Hypertension 1 , 292-298. BROUGHTON, A. & KORNER,P.I. 1983. Basal and maximal inotropic state in renal hypertensive dogs with cardiac hypertrophy. Am 3 Physiol 245, H33-H41. CROSS,C.E. 1962. Right ventricular pressure and coronary flow. Am 3 Physiol202, 12-16. DONALD,D.E. & ESSEX,H.E. 1954. Pressure studies after inactivation of the major portion of the canine right ventricle. Am 3 Physiol 176, 155-161. FOLKOW, B., HALLBACK, M., LUNDGREN, Y. & WEISS, L. 1970. Structurally based increase of flow resistance in spontaneously hypertensive rats. Acta Physiol Scand 79, 373-378. FUJISAWA, A,, SASAYAMA, S., TAKAHASHI, M., NAKAMURA, M., OHYAGI,A,, LEE, J-D., YUI, Y. & KAWAI,C. 1984. Enhancement of left ventricular contracility after opening of an arteriovenous fistula in dogs. Cardiovasc Res 18, 51-59. GREDITZER, H.G. 111 & FISCHER, V.W. 1978. A sequential ultrastructural study of different arteries in the hypertensive rat. Ex$ Mol Pathol29, 12-28. GREENSMITH, J.E. & DULING, B.R. 1984. Morphology of the constricted arteriolar wall : physiological implications. Am 3 Physiol 247, H687-H698. D.G., FLORENTINE, M.S., BROOKS, L.A., HARRISON, M.L. 1988. The effect of COOPER, S.M. & MARCUS, hypertension and left ventricular hypertrophy on the lower range of coronary autoregulation. Circulation 77, 1108-1115. HEYMANN, M.A., PAYNE,B.D., HOFFMAN, J.I.E. & RUDOLPH,A.M. 1977. Blood flow measurements with radionuclide-labeled particles. Prog Cardiovasc Dis 20, 55-79. S., MARUYAMA, Y., KOIWA, Y., ISHIDE,N., ISOYAMA, KITAOKA, S., TAMAKI, K., SATO,S., SHIMIZU, Y., INO-OKA,E. & TAKISHIMA, T . 1981. Experimental study of afterload-reducing therapy. The effects of the reduction of systemic vascular resistance on cardiac output, aortic pressure and coronary circulation in isolated, ejecting canine hearts. Cwculataon 64, 490-499.
516
3.3. Smolich
et al.
J.V., FITCH, W., MACKENZIE, E.T., STRAND- pressure reduction in the autonomically blocked dog : evidence for right ventricular autoregulation. S. & HARPER,A.M. 1976. Lower limit of Cardiovasc Re5 22, 17-24. cerebral blood flow autoregulation in experimental J.J., WEISSBERG, P.L., BROUGHTON, A. & renovascular hypertension in the baboon. Circ Res SMOLICH, KORNER,P.I. 1988b. Aortic pressure reduction 39, 555-557. redistributes transmural blood flow in dog left KORNER, P.I. & HILDER,R.G. 1974. Measurement of ventricle. A m 3 Physiol 254, H361-H368. cardiac output and regional blood flow by thermoG.W. & COCHRAN, W.G. 1980. Statistical dilution. Clin Exp Pharmacol Physiol Suppl 1, SNEDECOR, methods. 7th ed. Iowa State University Press, Iowa. 47-66. S. 1976. Autoregulation of cerebral STRANDGAARD, LEE,R.M.K.W., FORREST, J.B., GARFIELD, R.E. & blood flow in hypertensive patients. The modifying DANIEL,E.E. 1983. Comparison of blood vessel1 influence of prolonged antihypertensive treatment wall dimensions in normotensive and hypertensive on the tolerance to acute, drug-induced hyporats by histometric and morphometric methods. tension. Circulation 53, 720-727. Blood Vessels 20, 245-254. SUGA,H., HISANO, R., HIRATA, S., HAYASHI, T. & MARCUS, M.L. 1983 The coronary circulation in health NINOMIYA, I. 1982. Mechanism of higher oxygen and disease. McGraw-Hill, New York. consumption rate : pressure-loaded vs. volumeMILLER,M.E., CHRISTENSEN, G.C. & EVANS,H.E. loaded heart. A m 3 Physiol 242, H942-H948. 1964. Anatomy of the dog, pp 283-285. WB TAYLOR, R.R., CINGOLANI, H.E., GRAHAM, T.P. & Saunders, Philadelphia. CLANCY, R.L. 1967. Myocardial oxygen consumpMURRAY,P.A. & VATNER,S.F. 1980. Fractional tion, left ventricular fibre shortening and wall contributions of the right and left coronary arteries tension. Cardiovasc Res 1, 219-228. to perfusion of normal and hypertrophied right TOMANEK, R.J., WANGLER, R.D. & BAUER, C.A. 1985. Prevention of coronary vasodilator reserve decventricles of conscious dogs. Circ Res 47, 190-200. rement in spontaneously hypertensive rats. HyNAKANO, J. & DE SCHRYVER, C. 1964. Effects of pertension 7, 533-540. arteriovenous fistula on systemic and pulmonary URABE,Y., TOMOIKE, H., OHZONO,K., KOYANAGI, S. circulations. A m 3 Physiol 207, 1319-1324. & NAKAMURA, M. 1985. Role of afterload in NORESSON, E., HALLBACK, M. & HJALMARSSON, A. determining regional right ventricular performance 1977. Structural ‘resetting’ of the coronary vascular during coronary underperfusion in dogs. Circ Res, bed in spontaneously hypertensive rats. Acta Physiol 57, 9&104. Scand 101, 363-365. R. D., PETERS,K. G., MARCUS, M. L. & OWENS,G.K., SCHWARTZ, S.M. & MCCANNA,M. WANGLER, TOMANEK, R. J. 1982. Effects of duration and 1988. Evaluation of medial hypertrophy in reseverity of arterial hypertension and cardiac hypersistance vessels of spontaneously hypertensive rats. trophy on coronary vasodilator reserve. Circ Res Hypertension 11, 198-207. 51, 10-18. SARNOFF, S.J., BRAUNWALD, E., WELCH,G.H. JR., WICKER, P. & TARAZI, R. C. 1985. Right ventricular W.N. & MACRUZ, R. 1958. CASE,R.B., STAINSBY, coronary flow in arterial hypertension. A m Heart 3 Hemodynamic determinants of oxygen consump110, 845-850. tion of the heart with special reference to the YONEKURA, S., WATANABE, N., CAFFREY, J.L., GAUGL, tension-time index. A m 3 Physiol 192, 148-156. J.F. & DOWNEY, H.F. 1987. Mechanism of attenuSMOLICH, J.J., WEISSBERG, P.L., BROUGHTON, A. & ated pressure-flow autoregulation in right coronary KORNER, P.I. 1988a. Comparison of left and right circulation of dogs. Circ Res 60, 133-141. ventricular blood flow responses during arterial
JONES,
GAARD,
