| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 1996;61:1381-1387
© 1996 The Society of Thoracic Surgeons
Departments of Cardiothoracic Surgery and Anesthesiology, Medical College of Wisconsin, Milwaukee, Wisconsin
Accepted for publication January 17, 1996.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Methods. Swine (46 to 54 kg; n = 7) were anesthetized and instrumented to measure instantaneous RV pressure, septal-to-RV free wall diameter (SFWD), RV free wall segment length (FWSL), RV volume via conductance (CV), and pulmonary artery flow, the integral of which was used as the standard for stroke volume. Flow-derived stroke volume was correlated with the systolic change in CV, FWSL, and SFWD in the steady state after incremental volume loading and on a beat-to-beat basis during transient inferior vena caval occlusion. Contractility was altered by calcium and pentobarbital and assessed by preload recruitable stroke work (PRSW).
Results. Mean (± standard error of the mean) correlations (r) versus stroke volume during steady state conditions were 0.85 ± 0.04 for FWSL, 0.83 ± 0.04 for CV, and -0.04 ± 0.24 for SFWD. Mean r values versus stroke volume during caval occlusions were 0.83 ± 0.03 for FWSL, 0.85 ± 0.04 for CV, and -0.03 ± 0.31 for SFWD. Calcium increased mean PRSW slope compared with control using CV (20.3 ± 2.6 versus 16.1 ± 1.9 mm Hg; p < 0.05), and pentobarbital decreased mean PRSW slope compared with control using both CV and FWSL (11.3 ± 1.0 versus 16.1 ± 1.9 mm Hg, p < 0.05; and 11.9 ± 2.1 versus 26.1 ± 4.0 mm Hg, p < 0.05, respectively). There were no changes in PRSW slope with either calcium or pentobarbital using SFWD. The PRSW function was linear with both FWSL and CV but not with SFWD.
Conclusions. In the normal heart, both FWSL and CV, but not SFWD, accurately reflect relative instantaneous RV volume and are thus useful for determining RV contractility by pressure–volume (pressure–dimension) indices.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Pressure–volume analysis for assessment of right ventricular function has been hampered by difficulty measuring instantaneous right ventricular (RV) volume. This problem results from the fact that crescentic RV geometry confounds the use of many techniques established for estimation of left ventricular (LV) volume. Methods described for calculation of RV volume from multiple linear dimensions or from spatial relations of implanted markers are laborious and not easily reproduced [1–5]. Standard echocardiographic techniques are not suited for measuring instantaneous volumes because imaging is required in multiple planes [6–8], and methods for assessing RV function using instantaneous echo-derived short-axis areas have only recently been validated [9]. Finally, the reliability of absolute RV volume determination by the conductance method is unclear [10–13].
Despite limitations, some of these techniques may still have utility for pressure–volume analysis [12–12–17]. Pressure–volume indices are useful for detecting changes in contractility after interventions in a particular animal or patient but not for defining a ``normal'' contractile state that could be broadly applied as a standard. Determination of absolute volume does not appear to be essential, provided that the method yields consistent and reproducible approximations of changes in absolute volume during the cardiac cycle. Accordingly, the purpose of this investigation was to demonstrate that relatively simple sonomicrometry and conductance techniques could be used to reflect instantaneous RV volume and could therefore be used in pressure–dimension (or volume) analyses for measurement of changes in global right ventricular contractile function in vivo.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
The femoral vessels were exposed, and systemic arterial pressure was measured with a fluid-filled catheter and manometer (Gould, Oxnard, CA). The chest was opened by median sternotomy and the heart exposed. A micromanometer-tipped catheter (Millar Instruments, Inc, Houston, TX) was inserted into the right atrium and advanced into the RV for measurement of RV pressure. A cylindric ultrasonic dimension crystal (Triton Technology Inc, San Diego, CA), which emits ultrasonic waves in a wide distribution, was inserted into the interventricular septum, and a hemispheric crystal was attached to the epicardium of the RV free wall to measure instantaneous septal-to-RV free wall diameter (SFWD). The septal crystal was advanced into the midseptum after creation of a tract with a 16-gauge catheter just to the right of the left anterior descending coronary artery, and its position was verified by epicardial echocardiography. The free wall crystal was placed at the midpoint of both the apex-to-base and anterior-to-posterior dimensions, and its alignment with the septal crystal was adjusted to maximize the dimension signal. A second crystal pair was used to measure instantaneous anterior RV free wall segment length (FWSL). These crystals were placed intramyocardially, 1.5 cm apart, with their axis aligned perpendicular to the atrioventricular groove and the rightward crystal approximately 2 cm from the groove. Both crystal pairs were connected to a sonomicrometer (Crystal Biotech, Hopkinton, MA).
A conductance catheter (Webster Labs, Baldwin Park, CA) was inserted into the pulmonary artery and advanced retrograde across the pulmonic valve to the RV apex. The catheter had one electrode at its tip, six evenly spaced electrodes within the ventricle to create five intracardiac segments, and one electrode above the pulmonic valve. The catheter was connected to a custom-designed signal processor, which applies a constant current (20 µA RMS @5 kHz) across the first and eighth electrodes and measures conductance between the six adjacent intraventricular electrodes. Catheter position was confirmed by palpation, epicardial echocardiography, and on-line assessment of segmental volume signals. A 20-mm ultrasonic flow probe (Transonic Systems Inc, Ithaca, NY) was placed around the main pulmonary artery and connected to a transit-time flowmeter (T101; Transonic). A snare was placed around the inferior vena cava (IVC) to transiently alter preload. Hemodynamics were allowed to stabilize after instrumentation.
Protocol
Intravenous heparin (10,000 units) and 5% human albumin (500 mL) were administered. A sample of blood was withdrawn to measure its conductance, and an intravenous bolus of hypertonic saline solution (10 mL) was rapidly injected to determine parallel conductance. Control data were collected. A 750-mL phlebotomy was then performed by gravity drainage into an empty intravenous bag. Data were collected after phlebotomy was completed and after the blood was reinfused in three 250-mL increments. Data were then collected after administration of the positive inotrope calcium chloride (750 mg intravenously) and the negative inotrope pentobarbital (450 mg intravenously). The animal was allowed to stabilize after reinfusion of the final 250-mL blood aliquot, and contractility measured at this point was used as the control against which comparisons were made after calcium and pentobarbital administration. Each pig was euthanized by induction of ventricular fibrillation in the presence of general anesthesia, and the position of the instruments was verified.
All animals received humane care in compliance with the ``Guide for the Care and Use of Laboratory Animals'' prepared by the Institute of Laboratory Animal Resources and published by the National Institutes of Health (NIH publication 85-23, revised 1985).
Data Analysis
Analog data were collected on an eight-channel recorder (Grass Instruments, Quincy, MA), digitized at 200 Hz, and stored on computer disc for later analysis. Data were collected during steady-state conditions and during brief (less than 10 seconds to avoid reflex responses) periods of IVC occlusion before and after each intervention.
Instantaneous ventricular volume, V(t), was derived from the conductance signals as described by Baan and associates [18] using the following formula:
 |
where 
is a dimensionless constant (assumed to be equal to 1), L is the interelectrode distance (1.6 cm), 
b is the specific conductivity of the blood measured directly, G(t) is the sum of the conductances measured between the intraventricular pairs of adjacent electrodes, and Vc is a volume term to correct for parallel conductance caused by structures surrounding the right ventricle. Parallel conductance can be determined by acutely altering b with intravenous hypertonic saline solution, which changes the relationship between G and V in equation 1. Conductance is recorded during rapid saline solution injection, and end-systolic conductance is plotted as a function of end-diastolic conductance on a beat-to-beat basis. This relation is linear provided that ejection fraction and parallel conductance remain constant and its point of intersection with the line of identity (end-systolic conductance = end-diastolic conductance) equals parallel conductance.
Stroke volume (SV) was determined by integration of the pulmonary blood flow waveform during ejection. Stroke volume was compared with the systolic change in RV conductance volume (CV), FWSL, and SFWD during steady-state conditions after the volume interventions described above and on a beat-to-beat basis during IVC occlusion.
Global RV contractility was assessed by preload recruitable stroke work as described by Glower and colleagues [19]. Pressure–volume (or pressure–dimension) loops were generated simultaneously during IVC occlusions using CV, FWSL, and SFWD. Volume stroke work (SWv) and dimensional stroke work (SWd) were defined by the area bounded by the loop (as determined by integration) and were plotted on a beat-to-beat basis as a function of end-diastolic volume (EDV) or dimension (EDD). These data were fitted by linear regression analyses to the following equations:
 |
 |
where Mwv and Mwd are the slope of the regression for volume and dimension, respectively, and Vw and Dw are the x-axis intercepts for volume and dimension, respectively. A minimum of ten beats during an IVC occlusion was used to develop each relation. The linearity of this relation has been demonstrated to be independent of loading conditions, unlike the end-systolic pressure–volume (–dimension) relation, which may exhibit nonlinearity at extremes of afterload [20]. Both slope and intercept data were compared by analysis of variance for repeated measures after the last blood reinfusion and after calcium and pentobarbital were administered. Statistical differences (p < 0.05) were defined by the least significant difference test.
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
. There were no significant changes in heart rate during the course of the experiment. Mean arterial pressure, RV systolic pressure, RV end-diastolic pressure, and RV SV all decreased after removal of 750 mL of blood. Arterial pressure did not recover fully until addition of calcium, but other hemodynamics recovered after return of the second increment (500 mL total) of blood.
|
. The systolic excursion in SFWD tended to remain constant or increase during reduction of preload while CV, FWSL, and SV all decreased in a parallel fashion. The flow probe remained well coupled with the pulmonary artery and provided excellent waveforms in all seven experiments during steady-state conditions and during IVC occlusions. The family of pressure–volume (–dimension) loops obtained from consecutive cardiac cycles during an IVC occlusion are shown in Figure 1B. Loops generated from conductance and FWSL data behaved similarly during preload reduction, with decreasing end-systolic and end-diastolic volumes (dimensions), concomitant with decreasing end-systolic pressure. Loops generated from SFWD data demonstrated an appropriate decrease in end-systolic dimension during IVC occlusion, but end-diastolic dimension remained constant.
|
. The correlation coefficient (r) in this example is high for both FWSL and CV but is negative for SFWD. Correlation coefficients for all experiments are shown in Table 2. Volume interventions resulted in five data points (three degrees of freedom for correlation) in each experiment; an r value greater than 0.88 was therefore required for p to be less than 0.05. The r values for FWSL versus SV after volume interventions exceeded 0.8 in six experiments and exceeded 0.9 in four. Similarly, the r values for CV versus SV exceeded 0.8 in five experiments and exceeded 0.9 in two. Correlations of SFWD versus SV were much more variable and were negative in four experiments. Mean (± standard error of the mean) r values for correlations versus SV resulting from volume interventions were 0.85 ± 0.04 for FWSL, 0.83 ± 0.04 for CV, and -0.04 ± 0.24 for SFWD. Correlations developed during IVC occlusions were based on a minimum of ten beats (eight degrees of freedom); r values greater than 0.63 were therefore required for significance in these analyses. The correlations for both CV and FWSL were significant in each experiment, but again, the relation of SFWD versus SV was quite variable. Mean r values for correlations versus SV resulting from IVC occlusions were 0.83 ± 0.03 for FWSL, 0.85 ± 0.04 for CV, and -0.03 ± 0.31 for SFWD.
|
|
. The relation is linear with both CV and FWSL but not with SFWD. Mean contractility data are shown in Table 3. Pressure–volume (–dimension) relations responded to inotropic interventions in a predictable fashion using conductance and FWSL, but not with SFWD. Mean Mwd increased after calcium administration from 26.1 ± 4.0 to 34.3 ± 7.9 mm Hg (p = not significant) and decreased after pentobarbital administration to 11.9 ± 2.1 mm Hg (p < 0.05 versus calcium and control) using FWSL. Mean Mwv also increased with calcium administration from 16.1 ± 1.9 to 20.3 ± 2.6 (p < 0.05), and decreased with pentobarbital administration to 11.3 ± 1.0 (p < 0.05 versus calcium and control) using CV. There were no significant changes in Mwd with administration of calcium or pentobarbital using SFWD. There were no significant changes in the volume or dimension axis intercepts after calcium or pentobarbital administration with FWSL, CV or SFWD. The mean r values indicate that the linearity of this relation was consistently high using CV, was slightly less consistent using FWSL, and was quite inconsistent using SFWD.
|
|
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
). Both techniques reflected the expected changes in myocardial contractility after administration of positive and negative inotropic agents. Assessment of instantaneous RV volume has been hindered by crescentic chamber geometry, which prevents application of most methods used to calculate LV volume. Two-dimensional echocardiography accurately measures RV volume by shell subtraction [6–8] but is not applicable to instantaneous measurement because imaging is required in multiple, simultaneous planes. Continuous measurement of single plane dimension by echocardiography may reflect instantaneous RV chamber volume, as recently described by Oe and associates [9]. A shell subtraction method using sonomicrometry to calculate instantaneous RV volume from multiple linear dimensions has been validated, but the technique suffers from lack of simplicity and perhaps reproducibility [1, 2, 16]. Another method requires implantation of multiple, biventricular markers that are tracked with biplane fluoroscopy during the cardiac cycle [3–5]. The markers create a polyhedron that is divided into smaller component tetrahedrons, each of whose volumes are solved from the xyz coordinates of the markers. This technique might intuitively seem best able to account for the unusual geometry of the RV, but the calculated volumes have not been validated against measured SVs or ex vivo pressure–volume relations. Similar to shell subtraction, this method is extremely cumbersome and may have limited reproducibility.
Some of the techniques in current use may still allow assessment of RV function by load-independent indices despite considerable limitations. The utility of current measures of contractility, such as the linearized Frank-Starling relation (preload recruitable stroke work), lies in their ability to quantitate changes in function over time after specific interventions. No standard values of ``normal'' function are defined by these indices. Although ideal, it is not essential to measure absolute volume, provided that a method accurately and consistently reflects changes in instantaneous volume over multiple cardiac cycles.
The conductance method has been previously validated for measurement of instantaneous LV volume in both animals and humans [18, 21]. In brief, the technique is based on the premise that instantaneous chamber volume is proportional to the sum of time varying conductances along a multielectrode catheter that is positioned within the cavity and used to apply an electrical field. The conductance technique has proved useful for assessing LV function by analysis of pressure–volume relations [22] despite concerns regarding its ability to measure absolute LV volume [23].
The present study confirms that conductance can also be used for investigation of RV function. Solda and coworkers [11] previously reported that conductance correlates well with known RV volumes in isolated rabbit hearts and with stroke volumes measured in vivo by integration of pulmonary flow signals; however, they failed to perform simultaneous, beat-to-beat analysis of SV and conductance. Several other investigators have recently correlated RV conductance with RV SV on a beat-to-beat basis [12, 13].
A linear relation between conductance-derived SV and flow-derived SV was demonstrated in the present study, but it can be appreciated from Figure 2 that the slope of this relation was not equal to unity. The term in equation 1 corrects for inhomogeneity of the electric field and was assumed to equal 1. This term is defined in Figure 2 as the slope of the regression, and the y-intercept would equal the product of and Vc. The value for in the LV is usually reported in the range of 0.7 to 1.0. We found the range in the RV to be 0.191 to 1.61 (mean, 0.528) in 7 animals. These observations agree with those recently reported by Dickstein and associates [12], whose data appear to demonstrate a range of 0.2 to 1.4 for values. The assumption that is equal to unity in equation 1 requires a relatively homogenous electric field and minimal current leakage across the myocardium. These conditions are less likely to exist in the crescentic, thin-walled RV than in the ellipsoidal, thick-walled LV and limit the ability of the conductance technique to measure absolute RV volume. Despite this, the present data demonstrate that conductance consistently measures relative RV volume and support the conclusions of Dickstein and associates [12] that RV pressure–volume loops generated with a conductance catheter responded to changes in inotropic state in a predictable fashion.
The conductance catheter was inserted into the RV in a retrograde fashion across the pulmonic valve of open-chest animals in the present experiments. Woodard and colleagues [10] have suggested that transatrial catheter placement may induce error by allowing transduction of atrial volumes across the thin tricuspid valve and found that transpulmonic placement was associated with superior signal quality. Conductance is thus a simple and accurate technique for assessing instantaneous RV volume in open-chest preparations, but may have less utility for closed-chest, chronic experiments if a route of insertion across the pulmonic valve is required.
The present data also confirm the correlation between in vivo RV SV and the systolic excursion in FWSL previously reported by Morris and colleagues [14]. A linear relation of these two measures might not be predicted intuitively based on RV geometry. Feneley and associates [1] showed that this relation is in fact nonlinear in isolated hearts if LV volume is varied while RV volume is held constant. These investigators speculated, therefore, that changes in LV volume during the normal cardiac cycle would result in a nonlinear relation in vivo, but did not test this hypothesis as was done in the present study.
The current data also demonstrate that RV FWSL can be used in conjunction with pressure data to reflect global RV contractility by load-independent analysis. Again, this might not be predicted because RV structure and function are quite heterogeneous [24, 25]. It is important to recognize, therefore, that the relation between regional and global function may vary throughout the RV and that data obtained in one region cannot be extrapolated to other regions. Positioning of dimension crystals for measurement of FWSL must be precise and consistent from experiment to experiment. The validity of FWSL measures would also be affected by regional alterations such as ischemia, not only in the area containing the crystals but potentially in other areas, such as the interventricular septum. Use of FWSL appears well-suited, however, for studies in which the effects of experimental interventions would be global in nature.
Septal-to-RV free wall dimension was found to be an unreliable index of RV stroke volume in the present investigation. Feneley and associates [1] did not study this particular relation in their studies using shell subtraction to determine RV volume, nor was it validated by Dickstein and associates [17], who used SFWD to assess RV function in open-chest pigs. The present data suggest that end-diastolic SFWD remains relatively fixed over a wide range of preloads throughout the cardiac cycle (see Fig 1). This may indicate that shortening that occurs in this dimension as a result of free wall contraction is offset by the normal leftward systolic movement of the interventricular septum. Other investigators have used SFWD as a surrogate for RV volume in developing linear indices of RV contractile function [26, 27]. The present data do not support this, nor did these investigators attempt to correlate SFWD with SV in their studies.
A potential limitation of this investigation is the use of an ultrasonic flow probe as the standard for SV. Concern has been expressed regarding the ability of Transonic S-series probes to measure absolute cardiac output [28]. The manufacturer concedes that these probes may indeed underestimate flow depending on vessel curvature, probe position, and maintenance of coupling but maintains that the accuracy of the probe is linear on a beat-to-beat basis during the course of acute experiments provided that positioning and coupling remain constant (M. Sosa, Applications Specialist, Transonics Systems, Inc; personal communication). In each of the present experiments, the flow signal was evaluated qualitatively and was noted to maintain excellent coupling with the pulmonary artery. This coupling was accomplished by packing autologous blood clot in any potential space that could result in loss of contact between probe and artery.
In conclusion, RV conductance and FWSL can be used in the normal heart as estimates of instantaneous volume to derive pressure–volume (–dimension) indices of global function. In contrast, RV SFWD appears to be an unreliable correlate of chamber volume and should not be used with RV pressure to assess contractile state. Further investigation with these measures is required to determine their relationship to instantaneous RV volume in other settings.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Related Article
This article has been cited by other articles:
|
|
 |
|
  A. M. Sheikh, C. Barrett, N. Villamizar, O. Alzate, A. M. Valente, J. R. Herlong, D. Craig, A. Lodge, J. Lawson, C. Milano, et al. Right ventricular hypertrophy with early dysfunction: A proteomics study in a neonatal model. J. Thorac. Cardiovasc. Surg., May 1, 2009; 137(5): 1146 - 1153. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. A. Leather, R. Ama', C. Missant, S. Rex, F. E. Rademakers, and P. F. Wouters Longitudinal but not circumferential deformation reflects global contractile function in the right ventricle with open pericardium Am J Physiol Heart Circ Physiol, June 1, 2006; 290(6): H2369 - H2375. [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]
|
|
|
|
 |
|
 R. H. LOPES CARDOZO, P. STEENDIJK, J. BAAN, H. A. A. BROUWERS, M. DE VROOMEN, and F. VAN BEL Right Ventricular Function in Respiratory Distress Syndrome and Subsequent Partial Liquid Ventilation . Homeometric Autoregulation in the Right Ventricle of the Newborn Animal Am. J. Respir. Crit. Care Med., August 1, 2000; 162(2): 374 - 379. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Sugimoto, M. Okada, N. Ozaki, T. Kawahira, and M. Fukuoka Influence of functional tricuspid regurgitation on right ventricular function Ann. Thorac. Surg., December 1, 1998; 66(6): 2044 - 2050. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. M. Heerdt and M. L. Dickstein Assessment of Right Ventricular Function Seminars in Cardiothoracic and Vascular Anesthesia, September 1, 1997; 1(3): 215 - 224. [Abstract] [PDF]
|
|
|
|
|
|
J. G. Markley and A. C. Nicolosi Effects of Left Heart Assist on Geometry and Function of the Interventricular Septum Ann. Thorac. Surg., December 1, 1996; 62(6): 1752 - 1758. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
