| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 1996;62:1752-1758
© 1996 The Society of Thoracic Surgeons
Department of Cardiothoracic Surgery, The Medical College of Wisconsin, Milwaukee, Wisconsin
Accepted for publication June 20, 1996.
 |
Abstract |
|---|
 TopFootnotes Abstract IntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Methods. Domestic swine (50 kg, n = 10) were anesthetized and instrumented for collection of physiologic data and for open-chest LHA, which was accomplished by left atrial to subclavian artery bypass using a centrifugal pump. Both global and regional RV function data as well as two-dimensional echocardiographic data of septal geometry and function were collected at control levels and during both partial and full LHA. Short-axis echocardiographic images were obtained at the midventricular level and analyzed to quantify septal curvature (k; cm-1), systolic septal thickening (%), and systolic septal excursion (cm).
Results. Partial LHA had no effect on either septal geometry or function. Full LHA resulted in decreased diastolic septal curvature (k = 0.10 ± 0.07 versus 0.42 ± 0.06 at control; p < 0.05), reduced systolic septal thickening (0.27 ± 5.23 versus 29.32 ± 8.61 at control; p < 0.05), and reversed leftward systolic septal excursion (-0.29 ± 0.11 versus 0.11 ± 0.03 at control; p < 0.05). End-diastolic septal position was shifted leftward during full LHA compared with control, but was associated with rightward systolic motion of the left ventricular mass and septum as a unit. There were no changes in global or free-wall RV function during either partial LHA or full LHA compared with control.
Conclusions. Left heart assist results in marked changes in both geometry and function of the interventricular septum in normal hearts. These changes, however, do not appear to be associated with changes in either global or regional RV function. Evaluation of the septum with echocardiography may be helpful in defining strategies for clinical application of this technology.
|
Introduction |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
A precise understanding of normal septal function during LHA and the ability to distinguish normal function from true dysfunction are important for optimizing strategies for clinical application of these devices. Accordingly, the purpose of this study was to quantify geometry and function of the interventricular septum during normal circulation and during both partial LHA (pLHA) and full LHA (fLHA) in normal pigs using two-dimensional echocardiography. We hypothesized that LHA in normal hearts would result in significant changes in septal geometry and function, but not in RV function. Echocardiography was chosen for this study because it is ideally suited for both quantitative evaluation of regional function [10] and intraoperative assessment.
|
Material and Methods |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
The animals were treated with heparin (300 U/kg intravenously) and cannulated for LHA. A 38F catheter was inserted into the left atrium through the appendage for drainage to the pump, and an 18F cannula was inserted into the right subclavian artery for arterial return. The tip of the drainage catheter was positioned in the posterior atrium, adjacent to the orifices of the pulmonary veins and not across the mitral valve. Air was removed from the cannulas, and they were connected to a centrifugal pump (Biomedicus, Eden Prairie, MN).
Protocol
The animal was allowed to stabilize after instrumentation, and then control data were collected. Hemodynamic data were collected with the animal apneic both during a steady state and during brief inferior vena cava occlusion. Two-dimensional echocardiographic data were collected using a 3.0-MHz linear phased-array transducer connected to an Irex Meridian system (Johnson and Johnson, Inc, Ramsey, NJ). Epicardial images were collected through a balloon partially filled with saline. Short-axis images were obtained at the level of the midpapillary muscles and stored on videotape for later analysis.
Partial LHA, defined as half the value of PA flow during control conditions, was then instituted. Hemodynamic and echocardiographic data were collected after stabilization. Full LHA, defined as the level of flow that prevented aortic valve opening, was then instituted, and data were collected subsequent to stabilization. The animal was then sacrificed by inducing ventricular fibrillation in the presence of deep general anesthesia.
All animals received humane care in compliance with the "Principles of Laboratory Animal Care" formulated by the National Society for Animal Research and the "Guide for the Care and Use of Laboratory Animals" prepared by the National Academy of Sciences and published by the National Institutes of Health (NIH publication 85-23, revised 1985). This study was approved by the Medical College of Wisconsin Animal Care Committee.
Data Analysis
Hemodynamic data were digitized at 250 Hz per channel and stored directly on computer disk using a commercial data acquisition package (Codas, Akron, OH). Right ventricular stroke volume (SV) was determined by integrating the PA flow waveform on a beat to beat basis. The transseptal pressure gradient was calculated by subtracting right from left ventricular pressure at end-diastole and at peak systolic pressure. A positive gradient was therefore arbitrarily defined as left pressure greater than right pressure.
Regional function in the RV free wall was determined by analysis of instantaneous pressure-dimension relations using customized software. A family of pressure-dimension loops was generated during transient inferior vena cava occlusion. Stroke work (SW) was defined on a beat to beat basis as the area bounded by each loop and was calculated by integrating RV pressure with respect to dimension over each cardiac cycle: SW = 
P x dL. Contractility was defined by the regional preload recruitable stroke work (rPRSW) relation, as described by Glower and colleagues [11]. Stroke work was plotted as a function of end-diastolic dimension (EDD) and fitted to the formula: SW = Mw(EDD - Dw), where Dw = the dimension axis intercept, and the slope of this relation (Mw) varies directly with contractile state in a load-independent fashion. End-diastole was defined as occurring 40 milliseconds before peak positive dP/dt.
The relation between flow-derived RV SV and the systolic excursion in FWSL was also determined with each intervention. We [12] and others [13, 14] have demonstrated in normal hearts that RV SV is closely correlated with FWSL, such that rPRSW can be used to reflect changes in global RV function. We wanted to see, however, whether a correlation exists between RV FWSL and SV during LHA.
Stop frame echocardiographic images were analyzed at end-diastole and end-systole in each experimental condition. End-diastole was defined by the peak of the R wave on simultaneous electrocardiogram, and end-systole was defined by the smallest chamber size associated with the simultaneous T wave. Images were transferred from the video monitor to a computer (model LCII; Macintosh, Cupertino, CA) using tracing paper and an X-Y digitizing tablet (model UD1212; Wacom, Vancouver, WA). Digitized images were analyzed using commercially available software (Canvas; Daneba Software, Miami, FL).
The anterior and posterior points of attachment of the RV free wall to the septum were identified, and the septal axis was defined as the straight line connecting them (Fig 1
). The midpoint of the axis (a) defined the origin of a rectangular coordinate system, using the septal axis as the ordinate and a perpendicular line through a as the abcissa. Points on the LV side of the septal axis were assigned positive values, and points on the RV side of the septal axis were assigned negative values. Point r was defined as the intersection of the RV septal endocardium with the abcissa, and point c was defined as the geometric center of the LV cavity determined by computer. This coordinate system was used for image analyses.
|
t; %) was defined by the formula: t = [(tES - tED) ÷ tED] x 100. Curvature (k; cm-1) is defined as the inverse of the radius, r, for any known arc. Septal curvature was determined by fitting circles of a known radius to the right endocardial border of the septum. The circle of radius r that best fit the septum defined k = 1/r. A k value greater than zero was defined as concave to the LV, and k less than zero as concave to the RV. The septum becomes flat as k approaches zero or as r approaches infinity. The RV cross-sectional area (A) was measured at end-diastole and end-systole, and fractional shortening of RV cross-sectional area (A; %) was determined by the formula: A = [(AED - AES) ÷ AED] x 100. Comparisons of both hemodynamic and echocardiographic data were made at control levels, with pLHA, and with fLHA. All statistical analyses were performed with commercially available software (SigmaStat; Jandel Scientific, San Rafael, CA) using one-way analysis of variance for repeated measures and Student-Newman-Keuls method when significance was achieved (p < 0.05). By convention, all measurements are listed as mean ± standard error of the mean unless otherwise noted.
|
Results |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
. Full LHA resulted in a flattened aortic pressure tracing and decreased LV peak and end-diastolic pressures. There was no change in RV or PA pressures or in PA flow. Mean hemodynamic data are summarized in Table 1. Heart rate did not differ from control with either pLHA or fLHA; nor did peak RV pressure, RV end-diastolic pressure, or PA pressure. Peak LV pressure and LV end-diastolic pressure both varied inversely with the level of LHA, as did both end-systolic and end-diastolic transseptal pressure gradients. Mean end-diastolic transseptal pressure gradient ultimately became negative with fLHA. Right ventricular SV and cardiac output remained constant during LHA compared with control.
|
|
, and the regression of ED FWSL versus regional stroke work from these same data is shown in Figure 4. This relation was highly linear, with mean correlation coefficients of 0.95 ± 0.02, 0.97 ± 0.01, and 0.94 ± 0.02 for control, pLHA, and fLHA, respectively. Mean data for rPRSW are summarized in Table 2. Neither pLHA nor fLHA had any effect on slope (Mw) or dimension axis intercept (Dw) compared with control. The relation between SV and systolic excursion in FWSL was also highly linear (Table 3). Mean correlation coefficients for this relation were 0.86 ± 0.07 at control, 0.96 ± 0.01 for pLHA, and 0.93 ± 0.02 for fLHA. Based on these data, rPRSW was also used to reflect global RV contractile function and suggested that global RV function was not altered at either level of LHA compared with control.
|
|
|
|
. Mean echocardiographic data are summarized in Table 4. No change in RV cross-sectional area shortening fraction was observed with either level of LHA compared with control. Systolic septal thickening at control (t = 27.23% ± 8.29%) was unaltered by pLHA (29.32% ± 8.61%), but was nearly eliminated by fLHA (0.27% ± 5.23%; p < 0.006). Mean systolic septal excursion at control (0.11 ± 0.03 cm) was not affected by pLHA (0.16 ± 0.05 cm; p = not significant), but was reversed (motion toward the RV) during fLHA (-0.29 ± 0.11 cm; p < 0.002). Mean end-diastolic curvature at control (kED = 0.33 ± 0.02 cm-1) was not affected by pLHA (0.42 ± 0.06 cm-1; p = not significant), but was reduced during fLHA (0.10 ± 0.07 cm-1; p < 0.001). Full LV unloading resulted in rightward septal concavity in 5 animals. Despite the changes in kED, neither level of LHA resulted in any changes in kES compared with control.
|
|
). There was systolic shortening of rc at control, in conjunction with leftward systolic septal excursion. With fLHA, however, when the septum demonstrated rightward systolic motion, the distance rc remained constant throughout the cardiac cycle, suggesting parallel motion of the LV and septum. |
Comment |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
It has been demonstrated clearly in laboratory models that septal dysfunction significantly impairs global RV systolic function during LHA. In isolated hearts, LV unloading results in only mild reduction of RV systolic function, but LV volume unloading with septal paradox (created by wide incision of the LV free wall) is associated with marked impairment of RV function [4]. Studies in intact animals have demonstrated that LHA alone has a minimal effect on RV function, but LHA with septal ischemia results in significant RV depression [5]. Clinical studies using echocardiography also suggested that poor outcome with LHA is directly related to the degree of septal dysfunction [3].
The effect of LHA on RV function in the normal heart is not clear. Isolated heart preparations have demonstrated that LV pressure-volume unloading is associated with reduced RV systolic and diastolic pressures, an effect that is believed to be mediated by reduced transmission of transseptal forces (anatomic interaction) [4, 15]. It has been suggested that LHA may impair RV contractility in normal hearts through this mechanism. Although Moon and colleagues [6] have demonstrated reductions in load-independent indices of global RV contractility during LHA in normal hearts, Elbeery and associates [8] found no detrimental effect of LHA on RV contractile function. It has been observed that LHA is associated with a leftward shift of the septum, although the effect of septal position on RV function is also unclear. Moon and colleagues [6] maintained that this shift was responsible for the changes observed in RV contractility in their study. However, studies by Chow and Farrar [9] and by Olsen and co-workers [16] suggested that septal shift during LV unloading does not alter RV function. Furthermore, studies by Hendry and associates [17], in which septal position was controlled by an intraventricular balloon and by manipulation of loading conditions, suggested that septal position is not an important determinant of RV function during LHA.
A precise understanding of the changes induced by LHA in the geometry and function of the normal interventricular septum is important. Because septal dysfunction may be an elemental mechanism for RV failure during LHA, development of clinical strategies for optimal application of LHA would require that one be able to differentiate the changes that might occur in the normal septum from actual septal dysfunction.
Institution of fLHA in the present study resulted in decreased end-diastolic septal curvature, although it remained slightly concave toward the LV. However, end-diastolic transseptal pressure gradient became reversed during fLHA, with RV greater than LV pressure. End-diastolic septal curvature did not appear to correlate with transseptal pressure gradient in individual animals or as a group. Septal curvature at control was fixed in a relatively narrow range (0.3 to 0.4 cm-1) despite variable pressure gradients (-5 to 15 mm Hg). These data suggest that normal diastolic septal geometry may be determined in part by other factors, such as passive elastic recoil after contraction. Septal curvature during fLHA was less than 0.2 cm-1 in 7 cases and less than 0 in 5, but again did not correlate with pressure gradient. The septum became quite thickened in diastole during fLHA, a phenomenon that may have altered passive compliance properties and therefore resulted in relative insensitivity to pressure gradients in the physiologic range.
End-systolic septal curvature was not altered by LHA, despite large changes in peak systolic transseptal pressure gradient. It may be that the change in peak systolic gradient resulting from left atrial to subclavian artery bypass was simply not large enough to effect a change in systolic septal geometry. Alternatively, these results may indicate that factors other than pressure gradient (eg, fiber orientation or attachment) determine systolic septal geometry as well. Reduction or reversal of systolic septal curvature during LHA might indicate septal dysfunction and could be associated with RV dysfunction, although this relation remains to be determined by further investigation.
Left heart assist was associated with significant changes in systolic septal function. Systolic septal thickening was nearly abolished during fLHA. Both septal and LV free-wall diastolic thickness were dramatically increased during fLHA compared with control; that the septum did not thicken in systole may reflect maximal overlap of actin and myosin filaments in diastole in these unloaded hearts. Full LHA was also associated with a reversal of systolic septal motion, from a leftward to a rightward excursion. Qualitative review of real-time echocardiograms during fLHA suggested that the septum and LV mass moved to the right as a unit in systole. This was confirmed by quantitative analysis. At control, the RV septal endocardium remained to the right of the septal axis throughout the cardiac cycle, but moved toward it (leftward) during systole. This was true despite inherent septal thickening, which might have reduced apparent motion of this surface. During fLHA, however, the RV septal surface moved in a rightward direction during systole without septal thickening. Furthermore, the distance from the RV septal endocardium to the geometric center of the LV cavity (distance rc, see Fig 1) remained constant in systole during fLHA, suggesting a parallel motion of the septum and LV mass as a unit. The determinants of septal motion have been considered by Pearlman and colleagues [18], who studied septal motion using echocardiography in patients with various volume-loading lesions. They concluded that septal motion is determined by the end-diastolic position of the septum relative to the total cardiac mass, which includes the blood pool. The present data agree with this concept.
Regional function of the RV free wall, as determined by the rPRSW relation, was also not altered by LHA in this study, in agreement with the results of Chow and Farrar [9]. This differs from the results obtained by Moon and colleagues [19], who demonstrated a reduction in the slope of the end-systolic pressure-dimension relation in the sinus portion of the RV during LHA. The reasons for this difference in results are not clear; however, their experiments were conducted in closed-chest animals, and constraints of the chest wall may have affected RV free-wall function. The current study was performed in open-chest animals.
The use of a regional (free wall) index of contractility to reflect global RV function in normal hearts has been validated [12–14]. Regional RV function is known to be heterogeneous [20, 21], and therefore use of a regional index to reflect global RV function, particularly in the setting of LHA, must be done with caution. Placement of crystals must be consistent from experiment to experiment, and the relation of SV to the systolic excursion of regional dimension must be monitored carefully. In the present study, FWSL remained well correlated with SV on a beat to beat basis, both at control and during LHA. These results allow extrapolation of information regarding global contractility from regional data and lead to the conclusion that LHA had no effect on RV contractility. These results agree with those reported by Elberry and co-workers [8] and Hendry and associates [17]. There were also no changes in either RV SV or fractional shortening of RV cross-sectional area, although these are certainly load-dependent measures of ventricular function.
This study suggests that during LHA, the normal but minimally loaded ventricular septum provides a stable support against which the RV free wall contracts. The fact that it does not thicken in systole suggests that it does not contribute with intrinsic contractile force, although the reversed (rightward) systolic motion of the septum and LV mass as a unit may contribute to global RV function through inertial forces. This study also supports the concept of a leftward shift of the septum at end-diastole during fLHA, as suggested previously [6–9]. The end-diastolic position of the RV septal endocardium relative to the septal axis was shifted to the left compared with control (0.02 ± 0.09 versus -0.72 ± 0.10 cm, respectively; p < 0.05). The present data disagree, however, with the concept that this change in diastolic septal position results in impaired RV contractility.
There are several potential problems that could affect the conclusions of this study. First, the method of transferring echocardiographic images to the computer, which was adapted from a previous study of regional LV function reported by Nicolosi and Spotnitz [10], may be subject to error in transference of images either from the video monitor to tracing paper or from the tracing paper to the computer. All image transfer in this study was done by one person to eliminate the potential for interobserver differences. Second, extrapolation of regional data to assess global RV function may be problematic, as discussed earlier. We believe that this methodology has been validated adequately in this study and in our previous work. Finally, LHA was accomplished in this study using a centrifugal pump with left atrial, rather than LV, drainage. This may have influenced the extent of unloading achieved and therefore may not be applicable to other techniques of LHA.
In conclusion, LHA results in significant changes in the geometry and function of the normal interventricular septum. These changes do not appear to be associated with any impairment of global RV function. Echocardiographic techniques used in the present study are easily applicable to clinical, intraoperative evaluation of septal and RV function in patients requiring mechanical LV assistance. The mechanisms underlying RV dysfunction in association with LHA remain unclear and require further investigation.
|
Acknowledgments |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
|
Footnotes |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. E. Stone Current Status of Mechanical Circulatory Assistance Seminars in Cardiothoracic and Vascular Anesthesia, September 1, 2007; 11(3): 185 - 204. [Abstract] [PDF]
|
|
|
|
 |
|
 S Kant SEPTAL MOTION AFTER PDA CLOSURE Asian Cardiovasc Thorac Ann, February 1, 2006; 14(1): 90 - 90. [Full Text] [PDF]
|
|
|
|
 |
|
 B. P. J. Leeuwenburgh, W. A. Helbing, P. Steendijk, P. H. Schoof, and J. Baan Effects of acute left ventricular unloading on right ventricular function in normal and chronic right ventricular pressure-overloaded lambs J. Thorac. Cardiovasc. Surg., March 1, 2003; 125(3): 481 - 490. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. C. Nicolosi, G. West, J. G. Markley, B. Logan, and G. N. Olinger Gadolinium attenuates regional stunning in the canine heart in vivo J. Thorac. Cardiovasc. Surg., July 1, 2002; 124(1): 57 - 62. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. K. Karunanithi and M. P. Feneley Limitations of unidimensional indexes of right ventricular contractile function in conscious dogs J. Thorac. Cardiovasc. Surg., August 1, 2000; 120(2): 302 - 312. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. C. Nicolosi, J. G. Markley, and G. N. Olinger EFFECTS OF POSTISCHEMIC LEFT VENTRICULAR PRESSURE-VOLUME UNLOADINGON CONTRACTILE RECOVERY AND MYOCARDIAL BLOOD FLOW IN THE REGIONALLY STUNNEDCANINE HEART J. Thorac. Cardiovasc. Surg., July 1, 1999; 118(1): 181 - 188. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
