| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 1997;63:294-297
© 1997 The Society of Thoracic Surgeons
Division of Cardiac Surgery, Brigham Women's Hospital, Harvard Medical School, 75 Francis St, Boston, MA 02115
To the Editor:
We read with great interest the recent article by Kim and associates [1] regarding the effect of intraaortic balloon pumping (IABP) on characteristic arterial impedance (Zc) and systemic vascular resistance (SVR). In summary, they analyzed the physiologic effect of IABP placed in the 1:3 assist mode. In such a mode, the first cardiac cycle was termed the "unassisted" beat, the second the "augmented" beat with diastolic pressure augmentation, and the third beat the "assisted" beat. To determine the effect of IABP on the vasculature, they calculated and compared changes in Zc and SVR between the "unassisted" beat and the "assisted" beat (Figure 1 of the article). They reported a 24% reduction in SVR and a 21% reduction in Zc from the "unassisted" beats to the "assisted" beats in 25 patients. This resulted fundamentally from a measured decrease in average aortic pressure and an increase in stroke volume, comparing the assisted beat with the unassisted beat (Figure 1 of the article).
Although a decrease in SVR and Zc may partially account for these changes, we believe there are other issues that must be addressed before such a conclusion is made. Intraaortic balloon pumping has substantial effects on hemodynamics that must be considered before its effect on the vasculature can be assessed. In addition, the assumptions behind traditional formulas used to calculate Zc and SVR do not apply to the hemodynamic situation of IABP. Finally, we question the appropriateness of shifting aortic pressure, measured at the tip of the intraaortic balloon, to match aortic flow for impedance calculations.
The concept of counterpulsation originated with the classic work of Kantrowitz and Kantrowitz [2] in which coronary flow was augmented by delaying the arterial pressure pulse via an extracorporeal circuit. Later, Clauss and colleagues [3], in one of the first implementations of this idea, physically removed blood from the circulation during systole and returned it during diastole. Modern counterpulsation began with the deflation and inflation of a gas-driven balloon placed in the aorta [4]. Whether one is physically removing and replacing blood into the arterial system or deflating and inflating an intraaortic balloon, the physiologic effect is the same. Such a device forms a truly active, additional ventricle in series with the natural left ventricle (LV) [5] to augment and assist the natural ventricle.
Intraaortic balloon pumping inflates a previously deflated balloon in the aorta during diastole of the augmented beat. By doing so, IABP in effect displaces a volume of blood in the aorta. This displaced blood raises diastolic pressure because now that volume of blood must occupy other space in the aorta, be pushed distally through the arterioles, or travel retrograde through the coronary vasculature. Physiologically, diastolic pressure is augmented because more space (blood plus inflated air in the balloon) now occupies the arterial compliance. Furthermore, more blood is pushed across the systemic vascular resistance, also generating increased pressure. Unless one accounts for the displaced blood produced by balloon inflation, one cannot measure changes in diastolic pressure from unassisted beat to augmented beat and attribute them to changes in SVR, Zc, or arterial compliance. Thus, Kim and associates appropriately did not calculate changes in Zc and SVR from the unassisted beat to the augmented beat.
Instead, Kim and associates attempted to calculate changes in SVR and Zc from unassisted to assisted beats. However, the formulas used are based on two fundamental premises that are violated by IABP in the 1:3 mode. One is that the circulatory system is free of transient volume shifts. This assumption is essential in correctly calculating Zc and SVR from standard pressure and flow data, for it is only in the steady state that the ratio of the two quantities meaningfully describes the state of the vasculature. In a control state without IABP, equilibrium between injection of blood by the LV into the vasculature and its subsequent runoff is manifested when end-diastolic pressure does not rise or fall with subsequent beats. Although IABP timing parameters were not explicitly stated in the article, the balloon is typically deflated just before ejection, producing maximal systolic unloading [6]. In the 1:3 mode, deflation disturbs the equilibrium, producing the exact opposite physiologic effects of diastolic augmentation described above. Now blood in the aorta must suddenly fill the space previously occupied by the inflated balloon, decreasing pressure instantly because the arterial compliance now has a smaller volume filling it. Thus, the precipitous drop in pressure just before ejection is produced partially, if not solely, as a result of the volume deficit created by the collapsed balloon. Because the pressure at the beginning of ejection is decreased, mean aortic pressure will be decreased for the duration of the beat. Thus, calculating decreased SVR as an explanation for the cause of the decreased pressure is artifactual because the decreased pressure is the result of something else-the volume disturbance created by the balloon. Consider a more extreme example. If suddenly, during diastole, 1 L of blood was withdrawn from the circulation, then aortic pressure would drop drastically during systole of the next cardiac cycle. Certainly one would not take this decreased aortic pressure, calculate a decreased "SVR," and conclude that vascular tone is decreased. Other mechanisms are involved. Because the balloon is activated only once every three beats, steady-state equilibrium with the balloon, heart, and circulation does not occur during any individual beat. This is illustrated by noting that end-diastolic pressures of all three beats in Figure 1 are all different. On the other hand, if the 1:1 mode of IABP were used, equilibrium could be established, and one might then calculate SVR (although not with conventional formulas) as a true reflection of the vascular state with IABP.
Another fundamental assumption of the formulas used to calculate Zc and SVR is that the stroke volume measured at the aortic root is equivalent to the volume delivered per beat to the distal vasculature to create a pressure drop across the SVR. This may not be the case, however. In IABP, the volume deficit created by the deflated balloon is not trivial; it is generally 50 to 75 mL, which is similar to a typical human stroke volume. Depending on the exact timing of balloon deflation, less blood may very well be available to travel distally across the resistive arterioles during systole because some blood must now occupy the space previously taken up by the inflated balloon. (When does this blood then get to the circulation, one might ask? During diastole, when the balloon inflates and performs work on the circulation.) In other words, the effective stroke volume of blood injected into the distal circulation per beat during ejection may very well be less than the actual stroke volume injected into the aortic root.
Thus, one cannot simply measures pressures and flows in the aortic root and calculate, as Kim and associates did, Zc and SVR using formulas that do not account for the effect of the balloon itself. The calculation of Zc and SVR from pressures and flows measured at the aortic root assumes that nothing other than the injection of stroke volume by the LV is perturbing the system. It also assumes that there are no reflecting and interfering pressure and flow waves in the aorta. Such waves certainly occur when an IABP, placed at a distance away from the LV, is actively inflating and deflating. To accurately model IABP, the aorta must be represented by a "transmission line" analog model and the IABP and LV by flow or pressure sources separated by a finite distance [7]. This complex model is necessary to truly account for the "displaced volumes" and "volume deficits" caused by the IABP and to account for wave propagation and interference created by the LV and IABP. That interference has been shown to be critical to IABP performance [8]. Although it was alluded to in the article by the calculation of the "pulse wave reflection" coefficient, simply calculating this coefficient does not adequately describe wave interference and propagation.
Without accounting for these complexities, formulas using only ascending aortic flow and pressure for Zc and SVR are actually calculating driving point input impedance and its real component seen by the LV at the aortic root. Without wave interference or reflections, input impedance is equal to Zc, and its real component is equal to SVR. Both of these are fundamental properties of the vasculature per se. With wave interference and reflections, however, these calculated quantities are not Zc or SVR, but rather functions of Zc, the intraaortic balloon, the LV, and the timing ("the phase") between the LV and the balloon [8]. Thus, the reference by Goldfarb and associates [9] cited in the article as demonstrating reflexively decreased SVR actually does not. Two other references [10, 11] also cited as demonstrating decreased SVR, actually document decreased hind limb resistance, something very different from SVR.
Finally, we question the appropriateness of shifting aortic pressure, measured at the balloon tip, to match aortic flow, measured at the aortic root, in calculating input impedance. Impedance at a single point in the circulation is a complex quantity of the ratio between pressure and flow at that point, with the real component representing the magnitude of the ratio and the imaginary component representing the relative phase shift between pressure and flow. Even if the pressure wave morphology at the intraaortic balloon tip is exactly the same as that at the aortic root, shifting aortic pressure results in a loss of all phase shift information, which, as stated above, is critical in describing successful IABP [8]. Although such a method may be accurate where there is no appreciable wave interference, its use with IABP, where wave interference does occur, is problematic and needs to be justified.
References
Departments of Thoracic Cardiovascular Surgery, and Cardiology, Loyola University Medical Center, 2160 S First Ave, Maywood, IL 60153
To the Editor:
Chen and colleagues raise several issues regarding our study on the effects of intraaortic balloon counterpulsation (IABC) on arterial impedance [1]. Their major issue is that the decreased systemic vascular resistance (SVR) calculated from the pressure and flow in the ascending aorta in the assisted beat does not reflect a reduction of resistance in the peripheral vasculature. Their argument is that the increased stroke volume in the assisted beat (the cardiac beat after the balloon augmented beat at 1:3 pumping mode) is due to a drop in mean aortic pressure as a consequence of the blood volume displacement resulting from balloon inflation/deflation. All the patients of our study had a 40-mL adult-sized intraaortic balloon. The volume of blood displaced by the inflation of a balloon this size should be about 10 to 30 mL [2]. This blood volume displaced into the peripheral circulation should lead to a decrease in aortic pressure in the assisted beat and increase in stroke volume due to the decreased afterload. If the decreased arterial pressure in the assisted beats is due mainly to a balloon-mediated change in aortic blood volume, then there should exist a linear relationship between the change in the stroke volume and the change in aortic pressure. However, no statistically significant relationship between those two hemodynamic variables was observed in our study.
Chen and colleagues are correct in their assertion that SVR calculated from measurements of stroke volume and aortic pressure does not necessarily reflect the peripheral vascular resistance (PVR). To address this issue, we examined PVR in 6 patients who had undergone coronary artery bypass grafting with IABC in a 1:3 mode. Peripheral vascular resistance was determined from the volume flow rate of the dorsalis pedis artery (internal diameter of 1.4 ± 0.2 mm) measured by a 10-MHz linear array pulse wave Doppler velocimeter (ATL HDI-3000, Bothel, WA) divided by arterial pressure measured simultaneously in the radial artery. The value of mean perfusion pressure of the dorsalis pedis artery was assumed to be close to the value of the radial artery blood pressure. Volume flow rate in the assisted beat (0.51 ± 0.04 cm3/s) was not significantly different from the unassisted beat (0.49 ± 0.05 cm3/s). However, because of a lower arterial pressure in the assisted beat (64 ± 12 versus 73.5 ± 12.9 mm Hg; p = 0.0001), the PVR of the assisted beat (16,954 ± 743 dynes•s/cm5) was significantly lower than the PVR of the unassisted beat (19,064 ± 759 dynes•s/cm5; p = 0.0001). These results confirm our previously published findings of SVR differences between the assisted and unassisted beat [1].
Another important issue raised by Chen and colleagues is that arterial impedance should be determined with the IABC in the 1:1 mode, not in the 1:3 mode. The intraaortic balloon is an artificial ventricle that provides another pressure and flow source in series with the circulatory system. In addition, it produces an artificial rise in diastolic pressure, which is usually far greater than the natural peak systolic pressure. Therefore, a physiologic description of the effects of IABC on arterial impedance cannot be determined if it encompasses the period of balloon augmentation. Murthy and associates [3] developed a "transmission line" analog model to describe the effects of a deflated balloon on aortic impedance spectra. No calculations of arterial impedance were done with the balloon inflated. If aortic input impedance is determined in the balloon-augmented beat, this impedance represents the vascular impedance of the arterial system from the aortic root to the descending portion of thoracic aorta. It is not the total vascular hydraulic impedance working against the natural ejecting ventricle. Clearly, the inflation of the balloon creates a hemodynamic discontinuity between the left ventricle and the arterial system that confounds the calculation of arterial impedance variables. Our study showed that SVR calculated during the augmented beat was significantly increased due to the arterial pressure distortion caused by balloon inflation [1]. We have also found this distortion in measurements of PVR in the dorsalis pedis artery. The PVR of the augmented beat (206,244 ± 8,188 dynes•s/cm5; p = 0.0001) was significantly greater than the PVR of the preceding unassisted beat.
Chen and colleagues also suggest that impedance measurements can only be made during steady-state conditions. Nichols and Pepine [4] investigated the effects of abrupt changes in stroke volume and pressure on the input impedance spectra of single cardiac cycles in anesthetized dogs. Stroke volume was changed by introduction of an early premature ventricular contraction. Arterial input impedance remained constant despite significant beat-to-beat changes in stroke volume and pressure. The results of Nichols and Pepine support our measurements of impedance during single cardiac cycles in the absence of steady-state conditions [1].
Finally, Chen and colleagues questioned the appropriateness of shifting aortic pressure, measured at the balloon tip, to match aortic flow, measured at the aortic root, in calculating input impedance. A study with simultaneous pressure measurements of the aortic root and the proximal descending aorta (where the tip of the balloon is located) in humans showed that there was no significant difference in the contour and magnitude of the pressure between the two sites of the aorta [5]. This was confirmed by Fourier analysis of the pressures simultaneously obtained from the two aortic sites in 20 patients undergoing diagnostic catheterization in our laboratory (unpublished data). To use the proximal descending aortic pressure wave as an estimate of ascending aortic pressure, a phase shift was necessary to synchronize the onset of pressure and flow in our study. In addition, the validity of shifting proximal descending aortic pressure to the onset of aortic flow in determining input impedance was verified by input impedance measurements using a high-fidelity catheter with pressure and flow sensors mounted at the same location in 3 anesthetized dogs (unpublished data).
In summary, it is essential that the direct effects of IABC on arterial impedance must be evaluated by comparing the beat before (unassisted) and after (assisted) the balloon-augmented beat. Impedance measurements made during steady-state conditions in the 1:1 mode are confounded by distortions in arterial pressure that occur with each balloon inflation. In the 1:3 mode, the beat after balloon inflation is accompanied by a decrease in vascular resistance determined from both central and peripheral measurements of pressure and flow. The precise relationship between changes in SVR or PVR and peripheral arteriolar tone and tissue perfusion during IABC will require additional investigation.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
