Systolic Flow Augmentation in Hearts Ejecting Into a Model of Stiff Aging Vasculature
Influence on Myocardial Perfusion-Demand Balance
Abstract Age-related arterial stiffening and widening of the pulse pressure elevates ventricular systolic wall stress while it lowers diastolic coronary perfusion pressure. These changes are thought to adversely alter the balance between myocardial work load and blood supply. To test this hypothesis, the native compliant thoracic aorta was surgically bypassed by a stiff tube in reflex-blocked anesthetized dogs. Ventricular outflow was directed into either native aorta or the bypass; the latter resulting in an increase in arterial pulse pressure from 37.8 to 107.5 mm Hg (P<.001), with minimal change in mean pressure and flow. Cardiac work load was assessed by pressure-volume area (PVA), which combines external and internal left ventricular work and is linearly related to myocardial oxygen consumption (MV̇o2), and by MV̇o2 itself. Regional phasic and mean coronary flow were measured in the left anterior descending coronary artery, and global flow was assessed by radiolabeled microspheres. Myocardial supply-demand balance was assessed by comparing flow at matched PVA or MV̇o2, flow-PVA relations, and endocardial-to-epicardial flow ratios. When blood flow was directed into the stiff bypass tube, peak systolic pressure, wall stress, and PVA all rose nearly 50%, yet diastolic perfusion pressure fell by 20 mm Hg (all P<.01). Rather than being compromised, however, mean coronary flow rose by 34%, maintaining the same endocardial-to-epicardial flow ratio (≈1.1). Flow augmentation persisted when data were compared at matched work load (PVA or MV̇o2) and mean arterial pressure, as well as over a range of work loads (P<.001 from ANCOVA of flow-PVA relations). The increased flow resulted from enhanced systolic perfusion, which nearly equaled diastolic flow when ejection passed into the stiff bypass. These data counter the notion that cardiac coupling with a stiff arterial system (as with aging) necessarily compromises myocardial flow versus metabolic demand. However, the data highlight a greater role of systolic flow under such conditions and also raise the novel suggestion that enhanced pulsatility of the arterial pressure waveform may itself augment coronary perfusion.
Human aging results in lowered arterial compliance, which is largely due to a progressive degeneration of the elastic components within the walls of major proximal conduit vessels.1 2 As a consequence, the arterial pulse pressure widens, systolic hypertension develops, and ventricular wall stress and metabolic demand are increased while mean diastolic pressures may be reduced.3 4 5 6 7 8 Since most myocardial flow normally occurs during diastole, these abnormalities are thought to limit cardiac function by adversely influencing the myocardial supply-demand balance.6 7 9 However, such adverse cardiac effects of enhanced pulsatile load remain speculative.
To investigate this issue, we developed an experimental model in which the native thoracic aorta (NA) is surgically bypassed by a stiff plastic conduit linking the proximal ascending aorta to the subdiaphragmatic abdominal aorta.10 Cardiac outflow can be directed into either the compliant NA or the stiff bypass “aorta” (bypass tube [BT]). With flow through the BT, total peripheral resistance is little changed, whereas arterial compliance is markedly reduced. As a consequence, arterial pulse pressure increases nearly threefold at the same mean pressure and flow, and the resulting arterial pressure-flow and ventricular pressure-volume waveforms are physiological and quite similar to those observed in elderly individuals.10 In the present study, this preparation was used to determine whether increased pulsatile load due to vascular stiffening adversely alters the relation between cardiac work load and perfusion.
Materials and Methods
The thoracic aorta bypass preparation has been previously reported in detail.10 The protocol was reviewed and approved by the Animal Care and Use Committee of the Johns Hopkins University. Briefly, adult mongrel dogs (n=14) of either sex were anesthetized with pentobarbital (30 mg/kg IV) and ventilated on a volume respirator, along with enhanced inspired oxygen. Ventilation was guided by periodic checks of arterial blood gas levels, and adjustments were made accordingly to maintain a Po2 of ≥60 mm Hg. Both carotid and femoral arteries were cannulated by fluid-filled catheters to allow arterial blood sampling and pressure monitoring and to provide a route for cerebral perfusion after aortic bypass. A central venous cannula was placed for injection of hypertonic saline used for volume signal calibration (see below).
Animals underwent a midline sternotomy. The pericardium was opened, and the heart was suspended in a pericardial cradle. A micromanometer catheter (SPC 350, Millar Instruments) was passed through the mid left ventricular (LV) free wall to measure intracavity pressure and secured by a pursestring suture. An ultrasonic flow probe (No. 2R, Transonics) was placed around the left anterior descending coronary artery (LAD) (after the first major diagonal branch) for phasic and mean blood flow recording. The proximal ascending aorta was cleaned of fat and adventitia and then partially occluded with a C clamp at or just proximal to the base of the right brachiocephalic artery. This isolated a portion of the aortic wall onto which a vascular graft (Dacron, 1- to 1.5-cm internal diameter) was sewn by end-to-side anastomosis. Femoral arterial pressure was monitored to ensure adequate distal aortic flow during suturing. Next, the abdominal cavity was entered through a midline incision, an 8- to 10-cm portion of the aorta extending from the iliac bifurcation to the inferior mesenteric artery was isolated, and all side branches were ligated. A T tube cannula was inserted into the aorta at this site, abdominal flow was then reestablished, and the cavity was closed. The extravascular port of the T tube was linked with the ascending aorta graft by a plastic tube (50 cm long; internal diameter, 1 cm; Tygon), thereby completing the bypass. This tube had a linear compliance of 3×10−3 mL/mm Hg over the physiological pressure range, 1% of the NA compliance.
Additional instrumentation was as follows: Ventricular volume data were measured by a conductance catheter11 12 13 method (Sigma V, CardioDynamics), with the catheter inserted through the apex and placed so that its distal end was ≈1 cm above the aortic valve. All individual pressure-volume segments displaying counterclockwise motion (ie, intracavitary) were identified, and only these segments were combined to assess total volume. The catheter was secured by an apical pursestring suture. Proximal aortic pressure and flow were measured by a second ultrasonic flow probe placed between the aortic root and the proximal bypass graft anastomosis and by a micromanometer catheter introduced through the left brachiocephalic artery and secured at the same site. In four of the animals, an additional catheter was advanced into the coronary sinus. Blood was continuously extracted and compared with arterial blood to measure the arterial−venous oxygen difference (ΔAVo2) (AVOX Systems).
Once the preparation was completed, autonomic reflexes were blocked by hexamethonium chloride (10 mg/kg IV). The adequacy of blockade was assessed by testing for the absence of a heart rate response to varying preload or arterial pressure. Supplemental hexamethonium (5 to 10 mg/kg) was given if reflex activation recurred. After blockade, pharmacological support of contractility and blood pressure was needed, and this was provided by low-dose continuous infusion of epinephrine (1 to 3 μg · kg · min−1). The dose was titrated to achieve physiological arterial pressure and flow with NA perfusion. Once established, this dose was not altered during the experiment.
Cardiac output was directed into either the NA or BT by positioning occlusion clamps. Fig 1⇓ shows schematic diagrams for the two flow patterns. Placement of clamps at the proximal and distal BT anastomosis sites (Fig 1A⇓, sites a and b) directed blood only through the NA. When these clamps were repositioned on the NA just beyond the bypass anastomosis and at the diaphragm (Fig 1B⇓, sites c and d), flow passed only through the BT. Under this condition, nearly the entire intrathoracic aorta was excluded from the systemic arterial circulation; however, all other organs were perfused. Data were measured under each ejection condition after stable steady state hemodynamics were established (at least 10 minutes). Signals were digitized at 200 Hz, displayed, and recorded by use of custom-designed software.
In addition to these baseline data, measurements were made at varying levels of total cardiac work load. Work load was either altered transiently, by left atrial hemorrhage into a reservoir or by bi–vena caval occlusion (n=12), or at steady state, also achieved by hemorrhage (n=4). These steady state data were supplemented by results obtained from animals used in a previous study10 (n=9). Data recorded during these preload changes were used to assess coronary blood flow (CBF) at matched levels of total ventricular work load and myocardial oxygen consumption (MV̇o2) for both NA and BT conditions. They were also used to examine CBF–myocardial demand relations.
Endocardial-to-epicardial (endo/epi) flow ratios and verification of the regional LAD flow data were obtained by flow analysis using radiolabeled microspheres. Approximately 2×106 15-μm labeled spheres (New England Nuclear) were suspended in 10% dextran plus Tween, vortex-mixed for 20 minutes, and injected into the left atrium. Simultaneous withdrawal of peripheral arterial blood enabled regional flow calibration by standard technique.14 At the study conclusion, animals were euthanatized, the heart was removed and weighed, and the LV was divided into four or five slices each with 10 to 16 epicardial and endocardial segment pairs. The tissue was counted, and flows were expressed as milliliters per minute per gram. Reported microsphere flow data are the average results from all segments.
Volume Signal Calibration
Calibration of the conductance catheter required both slope and offset determination. The offset was estimated by the hypertonic saline method.12 13 Data were analyzed by methods of Baan et al14 and Lankford et al,12 and only concordant runs were used. Three to five separate estimates were averaged. The calibration slope was the ratio of stroke volume determined by flow probe to that of the uncalibrated catheter signal. This ratio was determined for each beat measured during transient or steady state preload reductions, and the results were averaged.10 Slope and offset calibration parameters were then applied to the volume signal to yield absolute chamber volume.
Verification of Flow Analysis
Myocardial perfusion was primarily assessed by the mid-LAD ultrasound flow signal. These flows were assumed to be in fixed proportion to total myocardial flow, determined anatomically by the ratio of mass perfused by the mid-LAD to total LV mass. These assumptions were tested in four animals in which regional ultrasound flow (in milliliters per minute) was directly compared with microsphere-derived total LV flow. Data were measured under varying conditions (NA versus BT, high versus low preload). Fig 2⇓ shows typical results from one dog, with a high correlation between the flow measurements. Group regression analysis yielded a similar result with R2=.98. The average ratio of regional to total flow from these data was 23±4% (similar to a 22% value reported previously15 ), with a coefficient of variation of 10%. This supports the assumptions underlying analysis of the mid-LAD flows.
Arterial vascular properties were measured from aortic pressure-flow data. Impedance spectra were determined by Fourier analysis of baseline steady state data under the two ejection conditions. Characteristic impedance was calculated as the average modulus from 5 to 12 Hz, and total resistance was calculated as the zero-frequency modulus. Peripheral resistance was calculated as the difference between mean resistance and characteristic impedance. Arterial compliance was estimated by the area method of Liu et al.16 A limitation of this or any method that assumes a simplified Windkessel model is that it implies no wave reflections. Reflections were present in the bypass preparation, and interpretations of compliance values must keep this limitation in mind. Nonetheless, the results provide a reasonable indication of overall compliance changes.17
Total ventricular work load was indexed by pressure-volume area (PVA) and MV̇o2. PVA is the sum of ventricular external and internal work and linearly correlates with MV̇o2 per beat.18 The PVA-MV̇o2 relation is fairly insensitive to changes in LV load,18 19 including reduced compliance generated by the present aortic bypass preparation.10 Fig 3⇓ displays an example of PVA determination and the relation between PVA and myocardial flow. Ventricular pressure-volume loops are shown at varying preloads, and these data define end-systolic and end-diastolic boundaries (Fig 3A⇓). End-systolic relations were often nonlinear and therefore fit to quadratic curves.20 The diastolic relation was fit by a monoexponential. At each preload, PVA was equal to the sum of stroke work and potential energy (PE), the latter equal to the area bounded by systolic and diastolic pressure-volume relations between end-systolic volume and Vo (the volume axis intercept of the end-systolic pressure-volume relation) (Fig 3B⇓). PE was determined by analytic integration of the pressure-volume curve fits. From these data, beats could be selected at the highest matched PVA for comparison. In addition, by combining several such beats, relations between PVA and CBF were obtained (Fig 3C⇓). Fig 3⇓ shows results from a transient LV preload reduction, and similar data were also obtained by using steady state preload change.
In a subset of animals, ventricular work load was directly assessed by MV̇o2. In four animals, this was determined by on-line ΔAVo2 analysis combined with mid-LAD flows. In an additional nine animals, total coronary sinus blood flow and ΔAVo2 were measured. Measurements were also made over a range of steady state work loads.
Ventricular function assessment included assessment of midwall fiber stress (ςf) and strain (εf) estimated by the method of Arts et al.21 Stress was calculated as follows: ςf=3Plv/ln(1+[Vw/Vlv]), where Plv and Vlv are ventricular pressure and volume, respectively, and Vw is chamber wall volume. Mean ejection and peak systolic stresses were calculated. Strain was estimated as follows: εf=1/3 · ln(1+3Vlv/Vw). Plots of ςf-εf loops during preload reduction yielded a highly linear end-systolic ςf-εf relation, from which end-systolic myocardial stiffness (slope) was derived. Since end-systolic pressure-volume relations were curvilinear in the majority of animals, contractile function was more easily assessed by the highly linear relation between stroke work and end-diastolic volume,20 as well as by the maximal derivative of pressure measured at the highest matched end-diastolic volume obtained for both vascular loading conditions. This last index assumed that dP/dtmax occurred before the onset of ejection in both NA and BT modes. This was true in all cases, with ejection starting an average of 12 to 16 milliseconds after dP/dtmax.
Myocardial supply-demand balance was assessed three ways. The first compared mean regional blood flow at the highest matched level of PVA or MV̇o2 for both NA and BT conditions. The second compared relations between CBF and PVA, determined from variable preload data as shown in Fig 3C⇑. A downward and/or rightward shift of this relation was interpreted as a fall in blood supply for any given demand level. Results were obtained under both transient and steady state loading conditions. The third method was based on the ratio of diastolic to systolic pressure-time integrals (DPTI/SPTI), as described by Buckberg and coworkers.22 23 SPTI is the area under the systolic portion of the arterial pressure wave, and DPTI is the difference between areas under diastolic portions of arterial and ventricular pressure waveforms, respectively.
Data are provided as mean±SD. Baseline conditions and data at matched PVA (or MV̇o2) were compared by paired t test. Statistical tests for an influence of ejection pattern (NA versus BT) on PVA–myocardial flow relations were made by ANCOVA for each heart. Combined data were analyzed by multivariate regression; dummy variables were used to code for interanimal variation about the mean.24 Statistical significance is reported at P<.05.
Vascular and Ventricular Responses
Arterial loading changes associated with directing flow through the BT were similar to those previously reported with this model.10 Peak aortic pressure increased and diastolic pressure decreased by +61.8 and −17.9 mm Hg, respectively, whereas mean aortic pressure and flow did not change significantly (Table 1⇓). Pulse pressure widened from 37.8±8.4 to 107.5±26.5 mm Hg (P<.001). Peripheral resistance and characteristic impedance were not significantly altered, likely reflecting the fact that both very proximal and distal components of the arterial tree were unchanged by the bypass. In contrast, estimated total arterial compliance was markedly reduced. The enhanced pulsatile load also increased LV metabolic demand (PVA) and peak and mean systolic wall stress (Table 1⇓). Systolic contractile function rose modestly, as assessed by the slope (stroke work–end-diastolic relation) whereas dP/dtmax measured at the same end-diastolic volume did not significantly change. End-systolic myocardial stiffness rose from 32.8±14.7 to 49.1±14.9 g · cm−2 · 10−2 (P<.05).
Fig 4⇓ shows representative examples of aortic pressure, flow, and corresponding pressure-volume loops for NA and BT conditions. The principal features of NA ejection were a narrow arterial pulse pressure, early rapid rise of systolic flow, and a square-shaped pressure-volume loop. For the BT, the pulse pressure widened markedly, peak flow diminished, and pressure-volume loop shape became trapezoidal, reflecting late-systolic pressure rise. A late rise in diastolic pressure due to reflected waves was often observed with BT. This would only tend to increase arterial compliance estimated by the area method,16 so actual compliance was likely even lower.
Despite the fall in mean diastolic aortic pressure, myocardial perfusion was enhanced when ejection was directed through the BT. Mid-LAD regional flow was 34.3±11.3 mL/min (corresponds to total LV perfusion of 1.2 mL min−1 · g−1) when ejection was directed into the NA. This rose to 45.9±15.6 mL/min when ejection passed through the BT. This 36% rise in mean flow measured in the LAD territory was virtually the same as the flow increase determined for the whole heart (35±25%) by use of radiolabeled microspheres.
To determine whether higher CBF simply reflected increased ventricular demand, data were analyzed at similar PVA (Table 2⇓) or MV̇o2 (Fig 5⇓). To match PVA, data at high preload for NA ejection were compared with results at slightly reduced preload for BT ejection by using beats obtained during transient load reduction. Even at matched PVA and at the same mean arterial pressure, mean flow was 21% higher with BT. This occurred despite a 25-mm Hg decline in mean aortic diastolic pressure.
Indexing demand by PVA assumed that it adequately reflected MV̇o2 and that this correspondence was similar for NA and BT conditions, as previously found.10 The results then predicted that ΔAVo2 should decline in order to have higher flow at the same metabolic demand. These assumptions and the prediction were directly tested in 13 additional hearts by comparing steady state cardiac cycles for both NA and BT conditions at similar levels of MV̇o2. The results displayed in Fig 5⇑ support the PVA-matched data, showing a 15% rise in mean CBF with BT ejection as well as a decline in ΔAVo2 (−10%) (both P<.05).
To further probe the nature of increased flow with BT ejection, phasic CBF patterns were systematically studied. This revealed that the mean flow increase was primarily due to higher flows during systole. Fig 6⇓ displays phasic CBF waveforms for two examples (including the animal shown in Fig 3⇑). When ejection passed through the NA, CBF was predominantly diastolic. In contrast, ejection into the stiff BT led to a marked rise in systolic flow nearly equaling that during diastole. By integrating flow during systolic ejection versus nonejection (diastole), the percentage of systolic CBF was determined (Table 2⇑). This averaged 25.8±9.2% for NA and 45.8±13.0% for flow via the BT (P<.01). One component of this enhanced flow related to lengthening of the systolic ejection period (from 41% to 48% of cycle length, P<.001); however, peak flow rates also rose markedly. Diastolic flow remained virtually unchanged despite a 20-mm Hg fall in mean diastolic pressure.
Enhanced flow at matched PVA was not restricted to the epicardium. Table 3⇓ provides endo/epi flow ratios measured in nine hearts at similar mean arterial pressure. The ratios were minimally different between NA and BT conditions. Previous studies have suggested that DPTI/SPTI could serve as an index of the adequacy of endocardial perfusion.22 23 The ratio is normally 1.2 to 1.5, and when it falls below 0.6, the endo/epi flow ratio also declines to ≤0.6, consistent with subendocardial ischemia.22 Interestingly, this earlier finding did not hold for the present data. Although the DPTI/SPTI ratio was 1.22±0.3 for NA flow and fell to 0.55±0.24 for BT flow, this did not result in a reduced endo/epi ratio or to contractile dysfunction suggestive of ischemia.
Perfusion–Work Load Relations
To test whether the rise in CBF at a single matched PVA or MV̇o2 persisted over a range of work loads, PVA-CBF (or MV̇o2-CBF) relations for NA and BT ejection were compared. For data obtained during transient preload reduction (eg, see Fig 3C⇑), ANCOVA revealed significant differences between the relations in 10 of 12 animals (Table 4⇓), with the relations shifting upward with BT. Fig 7A⇓ shows an example of these results. Since transient preload reduction may not have provided fully adequate time for CBF to adapt to the change in work load, data were also obtained using steady state changes in preload from a different group of animals.10 As shown by example in Fig 7B⇓, the results were similar. Both examples were supported by group analyses. Multivariate regression indicated a significant elevation in the CBF-PVA relation due to BT ejection. For the transient preload data, the regression equation was as follows: CBF=13.75+0.0094 · PVA+4.2 · GRP, where GRP, a categorical variable, is 0 for NA and 1 for BT (r=.977, P<.001). For the steady state data, the regression was as follows: CBF=36.0+0.036 · PVA+9.2 · GRP (r=.976, P<.001). The relative percent change in the offset due to BT ejection was similar for the two data sets. Differences in their slopes reflected use of regional LAD flow for transient data versus global flow for steady state data.
Direct comparisons between results measured during steady state and transient preload changes were obtained in two additional animals, with cardiac demand assessed by MV̇o2. MV̇o2 for transient load reduction assumed a constant ΔAVo2 equal to the value measured at initial rest volume. Both sets of data responded similarly to the switch from NA to BT vasculatures (Fig 8⇓) and were consistent with the preceding group analysis.
Potential Role of Enhanced Systolic Flow
The increased contribution of systolic CBF with BT ejection suggested that total coronary perfusion under this condition might be more sensitive to a given change in mean arterial pressure. To test this, we also compared data at a 40% lower PVA (Table 5⇓). At this load, mean coronary pressure was also reduced similarly for NA and BT data, although this was achieved with a disproportionate decline in systolic pressure for BT ejection. Interestingly, CBF fell 56% more with BT than with NA ejection. One component of this was a greater decline in diastolic flow, likely due to diastolic pressures falling below the autoregulating range. However, there was also a marked disparity in systolic flow reduction, which was nearly 100% greater with BT ejection. Thus, BT ejection increased the sensitivity of myocardial perfusion to changes in mean arterial pressure.
Vascular stiffening and systolic hypertension are important risk factors for increased morbidity and mortality from cardiovascular disease.2 6 25 26 Among the mechanisms for this association is thought to be an adverse coupling between the heart and arterial tree. Hearts ejecting into a low-compliance vascular system must generate a wide pulse pressure, simultaneously increasing systolic chamber wall stress while lowering diastolic coronary perfusion pressure. Since a dominance of diastolic coronary perfusion is assumed, this combination would be expected to compromise myocardial perfusion relative to demand. The present study tested this hypothesis and found, somewhat surprisingly, that coronary perfusion actually rose with greater pulsatile load despite the same mean pressure and total metabolic demand (measured by PVA or MV̇o2). This enhanced perfusion was due to a near doubling of systolic flow and maintenance of diastolic perfusion. Significantly, however, ejection into a stiff arterial load resulted in a greater sensitivity of perfusion to changes in mean arterial pressure. These results suggest that systolic pressures may play a greater role in overall myocardial perfusion in elderly individuals. They also provide novel evidence that pulsatile loading, perhaps from the shape of the arterial pressure waveform itself, can influence CBF.
Myocardial Flow-Demand Balance
In a prior study using a similar experimental model,10 we reported that imposition of an acute increase in pulsatile load neither adversely affected contractile function nor energetic efficiency. This seemed surprising, particularly if myocardial perfusion was compromised as a consequence of the disparity between systolic load and diastolic perfusion pressure. The present study provides insight into the mechanism for this preserved function, showing that flow is not compromised but often enhanced even at a similar work load and that the endo/epi flow ratio remains normal. Both findings argue against an imbalance between myocardial supply and demand. The lack of change in the endo/epi flow ratio is important for several reasons. It indicates that the increment in overall myocardial flow does not solely pertain to outer muscle layers. This makes it unlikely that endocardial flow versus work load was compromised. It also suggests that a substantial rise in systolic stress without simultaneous critical reduction in diastolic pressure, as is often found in elderly humans, is unlikely to compromise endocardial perfusion.
Interestingly, this result was not predicted by the analysis of DPTI/SPTI. Buckberg et al22 reported that canine hearts exposed to supra-aortic banding with rapid pacing or to an arteriovenous fistula had a decline in the DPTI/SPTI ratio from 1.5 to 0.5 along with a fall in the endo/epi flow ratio from 1.0 to 0.4. Although the marked reduction of arterial compliance in the present study produced a similar decline in DPTI/SPTI, there was no evidence of endocardial hypoperfusion. Differences in the interventions may explain this discrepancy. In the earlier study, supra-aortic constriction markedly increased left atrial pressure (near 40 mm Hg), suggesting that altered chamber diastolic loading may have compromised endocardial perfusion. In contrast, end-diastolic pressure was low and minimally altered by the switch between NA and BT ejection modes (Table 1⇑). For the arteriovenous fistula data, the aortic diastolic pressure was very low (40 mm Hg), and this likely contributed to reduced endocardial perfusion.
Our data are compatible with a recent chronic study reported by Watanabe et al.27 In this investigation, aortic compliance was reduced by bandaging the descending thoracic aorta and left brachiocephalic trunk with nylon tape. After 4 weeks, arterial systolic and diastolic pressure changes were modestly altered, but compliance fell from 0.5 to 0.24 mL/mm Hg, and mean flow measured in the left circumflex artery rose by 42%. However, these data were not obtained at matched work loads; therefore, some enhanced flow may have been due to greater metabolic demand. Our data clarify this observation, indicating that a component of the enhanced flow may not reflect higher work load.
Pulsatile Pressure and Coronary Flow
Myocardial flow is influenced by both cardiac energy requirements and coronary perfusion pressure. As reviewed by Feigl et al,28 these interdependencies can be represented on a three-dimensional surface, with a positive relation between CBF and MV̇o2 that itself shifts upward with increments in mean coronary perfusion pressure. To our knowledge, the present data are the first to suggest a fourth dimension, that of the arterial pulse pressure itself. At the same mean aortic perfusion pressure and PVA or MV̇o2, a wider pulse pressure appears capable of augmenting CBF. The relative role of mechanical or biochemical mechanisms to augmented flow remains unknown and is beyond the scope of the present investigation. However, it is noteworthy that studies have reported that nitric oxide plays an important role in flow-mediated endothelial relaxation in epicardial vessels29 30 and that pulsatile flow itself can further enhance this effect.31 It is conceivable that enhanced epicardial release of such substances influences downstream resistance vessels in an autocrine fashion, contributing to enhanced myocardial perfusion.
Increased pulsatile load also raised systolic wall stress, and this is traditionally anticipated to increase the systolic myocardial flow impediment particularly in the endocardium. However, recent studies have suggested that ventricular stiffness rather than stress may better define how myocardial properties dictate the balance between systolic and diastolic flow.32 33 34 These investigators propose that intramyocardial blood volume is pressurized during systole as a consequence of increasing myocardial stiffness and that this pressure counters forward coronary pressure to determine the extent of antegrade perfusion. A greater systolic stiffness would therefore impede flow more.
In the present experiments, both systolic stress and end-systolic stiffness increased when ejection was directed into the stiff BT. By either model, one would predict a greater systolic flow impediment, yet systolic flow nearly doubled. Thus, the systolic pressure augmentation was sufficient to overcome any increase in flow impediment. This raises a cautionary note in trying to predict how changes in arterial pulse pressure will influence the balance of the forces that generate forward and backward flow. What our results indicate is that even for exaggerated pulsatile loading, systolic perfusion pressure rise exceeds that of ventricular stress or stiffness, such that antegrade systolic perfusion is actually enhanced and not impeded.
The present study may have implications for management of arterial blood pressure in the elderly. The shift to a less compliant aorta resulted in a greater dependence on systolic flow, and with similar reductions of PVA and mean and diastolic coronary pressures, myocardial flow declined disproportionately. When blood pressure is clinically regulated, it is usually the mean and diastolic pressures that receive the greatest attention. Our data suggest that systolic pressure should also be carefully considered, because excessive lowering of systolic pressure may result in an acute adverse limitation of myocardial perfusion. This could be particularly important for hearts with reduced coronary flow reserve, which is also found with aging35 or with coronary stenoses. One important caveat is that the present data were measured in normal (not elderly) ventricles and with an acute rather than chronic imposition of a high pulsatile load. Chronic adaptations and changes in the aging myocardium and vasculature could alter these responses, and further studies will be needed.
The present study was conducted by using a preparation in which ventricular work load and coronary perfusion were directly coupled. This was necessary to determine whether enhanced pulsatile loading of the ventricle would indeed compromise coronary perfusion relative to energy demand. However, this meant that we could not test for independent influence of pulsatile coronary perfusion on coronary flow and that we could not test whether pulsatile pressures alter myocardial autoregulation. These are questions for ongoing investigations in which pulsatile coronary perfusion can be independently controlled.
Because of the complexity of the on-line measurements and the integrated physiology, it was nearly im- possible to match cardiac work load (either PVA or MV̇o2) for both NA and BT conditions at the time of study. Rather, cardiac cycles were obtained over a range of loads and from beats selected among these to obtain the highest similar work load. Preload change was used to vary demand, although this simultaneously lowered mean arterial perfusion pressure. Significantly, and as shown in Tables 2⇑ and 4⇑, mean coronary pressures were similar for the two conditions at matched work load and remained well above a potential ischemic range. The similarity of results obtained by transient compared with steady state preload changes further supports the dominance of metabolic demand over perfusion pressure in mediating the flow response. Transient loading changes might be expected to be more sensitive to changes in arterial pressure, yet results were qualitatively and quantitatively similar to those measured with steady state preload alteration. Prior studies have shown that CBF can respond quickly to changes in work load.36 If anything, flow may have lagged slightly behind changes in demand but would influence both NA and BT data.
We report that marked reductions in arterial compliance resulting in systolic hypertension, reduced diastolic pressures, and increased ventricular wall stress do not necessarily limit myocardial perfusion relative to total energy demand. On the contrary, in normal hearts with normal coronary vessels, myocardial flow can rise above that required by metabolic demand, largely because of the augmented flow during systole. This suggests that the pressure waveform itself may have an influence on myocardial flow. Finally, although arterial stiffening did not compromise myocardial supply-demand balance at rest, there was a greater sensitivity of flow to changes in mean arterial pressure. This may have important implications to the management of arterial pressure in the elderly.
This study was supported by Public Health Service grant HL-44545 (Dr Kass) and an American Heart Association, Maryland Affiliate, Inc, Fellowship Award (Dr Recchia). Dr Kass is an Established Investigator of the American Heart Association. Dr Saeki is a visiting fellow from Osaka (Japan) Medical College. The authors gratefully thank Richard Tunin for his excellent technical assistance in performing these studies.
- Received February 22, 1994.
- Accepted September 16, 1994.
- © 1995 American Heart Association, Inc.
Nicholas WW, O’Rourke MF. Aging, high blood pressure and disease in humans. In: McDonald’s Blood Flow in Arteries. 3rd ed. Baltimore, Md: Edward Arnold Publishers Ltd; 1990:398-420.
Merillon JP, Motte G, Masquet C, Azancot I, Guiomard A, Gourgon R. Relationship between physical properties of the arterial system and left ventricular performance in the course of aging and arterial hypertension. Eur Heart J. 1982;3:95-102.
Kelly R, Hayward C, Avolio A, O’Rourke M. Non-invasive determination of age-related changes in the human arterial pulse. Circulation. 1989;80:1652-1659.
Weisfeldt ML, Lakatta EG, Gerstenblith G. Aging and cardiac disease. In: Braunwald E, ed. Heart Disease. 4th ed. Philadelphia, Pa: WB Saunders Co; 1992.
Yin FCP. The aging vasculature and its effects on the heart. In: Weisfeldt ML, ed. The Aging Heart. New York, NY: Raven Press Publishers; 1980:137-213.
Lakatta EG, Mitchell JH, Pomerance A, Rowe G. Human aging: changes in structure and function. J Am Coll Cardiol. 1987;10:42A-47A.
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.
Kass DA, Yamazaki T, Burkhoff D, Maughan WL, Sagawa K. Determination of left ventricular end-systolic pressure-volume relationships by the conductance (volume) catheter technique. Circulation. 1986;73:586-595.
Lankford E, Kass DA, Maughan WL, Shoukas AA. Does the parallel conductance of the volume catheter vary with the cardiac cycle? Am J Physiol. 1990;258(Heart Circ Physiol 27): H1933-H1042.
Baan J, van der Velde ET, de Bruin HG, Smeenk GJ, Koops J, van Duk AD, Temmerman D, Senden J, Buis B. Continuous measurement of left ventricular volume in animals and humans by conductance catheter. Circulation. 1984;70:812-823.
Becker LC, Schuster EH, Jugdutt BI, Hutchins GM, Bulkley BH. Relationship between myocardial infarct size and occluded bed size in the dog: difference between left anterior descending and circumflex coronary artery occlusions. Circulation. 1983;67:549-557.
Liu Z, Brin K, Yin FCP. Estimation of total arterial compliance: an improved method and evaluation of current methods. Am J Physiol. 1983;252:H588-H600.
Ting C-T, Yang T-M, Chen J-W, Chang M-S, Yin FCP. Arterial hemodynamics in human hypertension: effects of angiotensin converting enzyme inhibition. Hypertension. 1993;22:839-846.
Suga H. Ventricular energetics. Physiol Rev.. 1990;70:247-277.
Suga H, Hisano R, Hirata S, Hayashi T, Ninomiya I. Mechanism of higher oxygen consumption rate: pressure-loaded vs. volume-loaded heart. Am J Physiol. 1982;242(Heart Circ Physiol 13):H942-H948.
Kass DA, Beyar R, Lankford E, Heard M, Maughan WL, Sagawa K. Influence of contractile state on curvilinearity of in situ end-systolic pressure-volume relations. Circulation. 1989;79:167-178.
Buckberg GD, Fixler DE, Archie JP, Hoffman JIE. Experimental subendocardial ischemia in dog with normal coronary arteries. Circ Res. 1972;30:67-81.
Brazier J, Cooper N, Buckberg G. The adequacy of subendocardial oxygen delivery: the interaction of determinants of flow, arterial oxygen content and myocardial oxygen need. Circulation. 1974;49:968-977.
Slinker BK, Glantz SA. Multiple linear regression is a useful alternative to traditional analysis of variance. Am J Physiol. 1988;255:H588-H600.
Kannel WB, Dawber TR, McGee DL. Perspectives on systolic hypertension: the Framingham study. Circulation. 1980;61:1179-1182.
Chu A, Chamber DE, Lin CC, Kuehl WD, Cobb FR. Nitric oxide modulates epicardial coronary basal vasomotor tone in awake dogs. Am J Physiol. 1990;258(Heart Circ Physiol 27):H1250-H1254.
Chu A, Chamber DE, Lin CC, Kuehl WD, Palmer RMJ, Moncade S, Cobb FR. Effects of inhibition of nitric oxide formation on basal vasomotion and endothelium-dependent response of the coronary arteries in awake dogs. J Clin Invest. 1991;87:1964-1968.
Rubanyi GM, Romero JC, Vanhoutte PM. Flow-induced release of endothelium-derived relaxing factor. Am J Physiol. 1986;250(Heart Circ Physiol 19):H1145-H1149.
Krams R, Sipkema P, Zegers J, Westerhof N. Contractility is the main determinant of coronary systolic flow impediment. Am J Physiol. 1989;257:H1936-H1944.
Krams R, Sipkema P, Westerhof N. Varying elastance concept may explain coronary systolic flow impediment. Am J Physiol. 1989;257:H1471-H1479.
Resar JR, Livingston JZ, Yin FCP. In-plane myocardial wall stress is not the primary determinant of coronary systolic flow impediment: a study in the isolated, perfused dog septum. Circ Res. 1992;70:583-592.
Egashira K, Inou T, Hirooka Y, Kai H, Sugimachi M, Suzuki S, Kuga T, Urabe Y, Takeshita A. Effects of age on endothelium-dependent vasodilation of resistance coronary artery by acetylcholine in humans. Circulation. 1993;88:77-81.
Schwartz GG, McHale PA, Greenfield JC. Coronary vasodilation after a single ventricular extra-activation in conscious dog. Circ Res. 1982;50:38-46.