Ann Thorac Surg 1996;61:1199-1204
© 1996 The Society of Thoracic Surgeons
Original Article: Cardiovascular
Right Ventricular Blood Flow During Left Ventricular Support in an Experimental Porcine Model
PaulJ. Hendry, MD,
Howard Nathan, MD,
Kesava Rajagopalan,
BEng
Divisions of Cardiac Surgery and Anesthesia, University of Ottawa Heart Institute, Ottawa Civic Hospital, Ottawa, Ontario, Canada
Accepted for publication December 15, 1995.
 |
Abstract
|
|---|
Background. Right ventricular blood flow may be adversely affected during left ventricular assist device (LVAD) use leading to right ventricular (RV) ischemia and RV dysfunction. This study characterized normal RV blood flow responses to LVAD operation.
Methods. Seven Yorkshire pigs weighing 74.4 ± 3.4 kg underwent right coronary artery blood flow measurements with an ultrasonic flow probe and injection of radiolabeled microspheres. A Thoratec LVAD was used in either synchronous or asynchronous modes and RV loading was increased using a pulmonary artery snare.
Results. The RV blood flow was compared between three regions that differed in proximity to the right coronary artery: proximal segment, mid-RV, and distal. The right ventricular distal flow was 0.93 ± 0.07 mL • min-1 • g-1 compared with 0.74 ± 0.06 mL • min-1 • g-1 at right ventricular proximal flow during control measurements (p = 0.0001). This difference was maintained during LVAD operation in either synchronous or asynchronous modes and also during pulmonary artery constriction.
Conclusions. Global RV flow is not adversely affected by LVAD use. A flow gradient occurs along the right coronary artery with the distal vascular bed having relatively less reserve, which may be more susceptible to ischemia in patients with preexisting coronary disease or RV distention during LVAD use.
|
Introduction
|
|---|
Left ventricular assist devices (LVAD) are gaining acceptance for both temporary and long-term treatment of end-stage heart failure [1]. As clinical experience with LVADs has increased, it has become clear that right ventricular dysfunction is a major problem. At centers where biventricular assistance is an option, some degree of right ventricular (RV) dysfunction necessitating biventricular assist device use is seen in 67% to 78% of patients [2, 3]. In other centers where only univentricular devices are available, serious RV dysfunction after ventricular assist device implantation is identified in 19% of patients in whom mortality is 78% [4–6]. Right ventricular dysfunction does not develop in all patients; therefore, attention has to be paid to the mechanisms that result in some patients having RV failure once they are supported with an LVAD.
One possible mechanism for RV failure may be an alteration in myocardial oxygen supply/demand during LVAD operation. A change in myocardial oxygen supply from decreased right coronary artery (RCA) blood flow leading to ischemia is known to cause abnormalities in RV free wall motion and decreased contractility of the posterior interventricular septum both of which can reduce right ventricular output [7, 8]. Coronary artery disease is seen in 42% to 51% of LVAD patients [2, 6]; therefore, impaired RCA blood flow may be partially responsible for RV dysfunction in these patients [1, 5, 9, 10].
A change in myocardial oxygen demand can also occur with LVAD use. A leftward shift of the interventricular septum occurs during LVAD operation as the left ventricle is unloaded [9, 11, 12]. In patients in whom pulmonary artery pressures are not reduced during LVAD unloading, an increase in internal RV dimension may lead to an increase in RV wall stress resulting in increased myocardial oxygen demand. Both the possibilities of decreased supply and increased demand during LVAD use may be sufficient to lead to RV dysfunction.
Changes in RCA phasic blood flow in the presence of an LVAD have not been well described. Before exploring alterations in RV myocardial oxygen supply/demand as a mechanism of RV failure, we wanted to document the normal responses of the RV vascular bed during LVAD use in normal swine hearts. This study was designed to assess changes in global and regional blood flow of the RV during the use of an LVAD in two different modes: pumping synchronously with the native heart or asynchronously in a full-fill/full-eject mode.
|
Material and Methods
|
|---|
The experiments were approved by and performed according to guidelines established by the University of Ottawa Animal Care Committee. Seven Yorkshire swine weighing between 48 and 87 kg were anesthetized with ketamine (11 mg/kg), chloralose (60 mg/kg bolus, 15 mg • kg-1 • h-1 infusion) and inhaled halothane (2% to 5%), intubated and ventilated with room air and oxygen at 3 L/min. The right jugular vein was cannulated with a Swan-Ganz catheter for measurement of right atrial, RV, and pulmonary artery (PA) pressures and the right internal carotid artery was cannulated to measure systemic arterial pressure. A median sternotomy was performed and pericardial cradle created followed by insertion of a cannula in the left inferior pulmonary vein to allow measurement of left atrial pressure and injection of microspheres. An ultrasonic flow probe (Transonic Systems Corp, Ithaca, NY) was placed around the PA. A snare was placed distal to the flow probe to allow control of RV afterload. A smaller flow probe was placed around the proximal RCA to assess RCA flow. Regional RV myocardial blood flow was determined by injecting radiolabeled microspheres (57Co, 103Sn, 103Ru, 95Nb, and 46Sc), the order of which was randomized. Before manipulation of the heart, lidocaine was applied topically to the heart and injected intravenously to decrease ventricular dysrhythmias.
The right atrial, RV, PA, left atrial, arterial blood pressures, heart rate (HR, by surface electrocardiograms), PA flow (cardiac output), and RCA flow were digitized on a computer at 200 Hz for later analysis. Stroke volume (SV) was calculated as SV (mL) = RV cardiac output/HR and right ventricular stroke work (RVSW)(joules) = SV x (RVPsys - RVPdias) x 0.0136, and RV dP/dt (using the first differential of the digitized RVP tracing) (where RVPsys and RVPdias are right ventricular pressure systolic and diastolic, respectively). The right ventricular rate–pressure product was calculated as HR x RVPsys as an indicator of RV oxygen demand.
After control hemodynamic measurements and injection of the first set of microspheres, the animal was heparinized (300 µm/kg) and a 14-mm diameter cannula was grafted to the ascending aorta. A 51F apical cannula was inserted into the left ventricular apex. The cannulas were attached to a Thoratec Ventricular Assist Device (Thoratec Laboratories, Berkeley, CA). The ventricular assist device operates in one of two modes: synchronous (SYN) in which the pump ejects after a sensed QRS complex and asynchronous (ASYN) in which the pump ejects when the pump's blood sac is completely filled. After a 15-minute stabilization period, the device was turned on either in 2:1 synchrony with the electrocardiogram (SYN) or asynchronously in full-fill/full-eject mode (ASYN). The animal was allowed to stabilize, then hemodynamic measurements were taken and another set of microspheres was injected. The device was changed to the other mode, allowed to stabilize followed by another set of measurements. The starting mode was determined randomly.
Using a PA snare, the PA was slowly constricted until the RVP and pulmonary arterial pressure increased by 20% to 24% more than previous values while on the LVAD. The device was operated in either SYN or ASYN mode, allowed to stabilize before hemodynamic measurements and injection of microspheres. The PA snare was released to allow the animal to recover and then reapplied with the device operating in the other mode. Measurements were taken once again. Thus, there were five individual time points: control, ASYN and SYN modes, and ASYN and SYN modes during PA constriction.
After the experiment, the heart was excised and the RV free wall, septum, and LV free wall were dissected. Myocardial blood flow (microspheres) was determined in samples of the RV free wall representing areas of differing proximity to the RCA: the proximal segment (closest to the RCA), mid-RV, and a distal segment (furthest away from the RCA and adjacent to the septum); the interventricular septum including the apical and basal segments; and the LV free wall including the apex, mid, and basal segments (Fig 1
).
View larger version (37K):
[in this window]
[in a new window]
|
Fig 1. . Areas of right ventricle that were sampled for regional blood flow assessment: 1 indicates the area closest to the right coronary artery (proximal segment), 2 is mid-right ventricle (RV) free wall, and 3 is the area furthest away from the right coronary artery (distal segment).
|
|
Statistical Analysis
Digitized data were analyzed to assess mean pressures and flows. The RCA flow curves were divided into the native heart's systole (end-systole = end-ejection when the PA flow curve passes through 0) and diastole (end-diastole = inflection of the RVP curve and/or the peak of the R-wave on the electrocardiogram. The RCA flow was also assessed as a function of the operation of the LVAD with end-systole defined as end-ejection with the dichrotic notch of the aortic pressure tracing and end-diastole as the inflection of the aortic pressure curve.
Data were compared using a Wilcoxon rank sum test for hemodynamic variables with a p value of less than 0.05 being considered as indicating statistical significance. A multiple analysis of variance test was used to compare regional blood flow data. Also, paired t tests were used to compare continuous variables. For the purposes of this analysis, the manipulation at each of the five time points was considered a grouping variable.
|
Results
|
|---|
Seven animals weighing 74.4 ± 3.4 kg were used for the experiments.
Effect of Mode on Hemodynamics and Blood Flow
Hemodynamic parameters are shown in Table 1. Control values were similar to those seen in previous experiments [12] and there was a decrease in cardiac output when the LVAD was turned on in either mode compared with control. This difference was statistically significant only for the ASYN mode (p = 0.05). Mean aortic blood pressure decreased to 62% to 72% of control (p = 0.0008) with LVAD use regardless of mode of LVAD operation. Right and left atrial pressures remained constant regardless of mode. RV dP/dt, RVSW, and RV rate–pressure product in both LVAD modes were somewhat decreased compared with control but the differences were not statistically significant.
The RCA blood flow was determined by an ultrasonic flow probe and was assessed at control with systole timed off the PA flow curve and diastole timed off the RV pressure curve (as described in the Material and Methods section) (Table 2). After the LVAD was turned on, cardiac cycle time points were determined relative to the LVAD. Under control conditions, phasic RCA flow was greater in diastole (63%). During LVAD operation, most RCA flow occurred during the device's systole (78%). However, total blood flow (mL/min) was similar for all groups. Representative graphs depicting aortic pressure and RCA blood flows during control, LVAD ASYN mode with and without PA constriction are shown in Figures 2, 3 and 4.
View larger version (33K):
[in this window]
[in a new window]
|
Fig 2. . Aortic blood pressure (AOP) (mm Hg) and right coronary artery blood flow (COQ) (mL/min) during control measurements.
|
|
View larger version (31K):
[in this window]
[in a new window]
|
Fig 3. . Aortic blood pressure (AOP) (mm Hg) and right coronary artery blood flow (COQ) (mL/min) during left ventricular assist device operation in asynchronous mode.
|
|
View larger version (29K):
[in this window]
[in a new window]
|
Fig 4. . Aortic blood pressure (AOP) (mm Hg) and right coronary artery blood flow (COQ) (mL/min) during left ventricular assist device operation in asynchronous mode during pulmonary aterial constriction.
|
|
Regional ventricular flows are shown in Table 2
. The left ventricular (LV) flow was quite uniform for all areas, therefore only a representative sample from the LV mid-myocardial region is shown for comparison. When RV and LV samples are analyzed together, the RV regions displayed a greater range of blood flows compared with a fairly narrow range of blood flows in the LV. The range of blood flow in the RV at control was 0.74 mL • min-1 • g-1 in the RV closest to the RCA (proximal) to 0.93 mL • min-1 • g-1 in the RV furthest from the RCA (distal). This regional flow difference was uniform in both ASYN and SYN modes as well as during PA constriction in both modes. If the RV regions alone are compared, there was greater flow distally compared with the proximal region (p = 0.0001), whereas in the LV the flow was relatively constant between regions (ie, 0.9 to 0.87 mL • min-1 • g-1 from base to apex). The RV regional flows were different from the LV regional flows (p = 0.002).
Effect of Pulmonary Arterial Constriction on Hemodynamics and Blood Flow
Heart rate and aortic blood pressure were similar to those measured without PA constriction. Mean RV pressures increased by 35% to 53% during PA constriction compared with control (p = 0.003). Cardiac output decreased to 43% to 56% of control with PA constriction (p = 0.0003). The RV stroke work and dP/dt were similar during all manipulations. The RV rate–pressure products were greater than those measured without PA constriction, but the difference was only statistically significant in the ASYN mode (p = 0.03).
With PA constriction, RCA blood flow was similar to values before constriction but increased compared with control. However, regional flows showed greater changes especially in their abilities to augment flow. Flow in the proximal region increased by 64%, whereas in the distal area it increased by only 41%. Although there was no statistically significant difference in RV regional blood flows between each time point (control, ASYN, SYN modes with or without PA constriction [p = 0.48]), there was uniformly a greater blood flow in the distal region compared with the proximal RV at each time point (p = 0.06).
|
Comment
|
|---|
This study showed that global RCA blood flow is not adversely affected by LVAD use. However, when assessing regional RV blood flows, there appears to be a difference in flows relative to distance from the RCA and a greater reserve to augment flow during stress in regions closer to the RCA.
Right ventricular function seems to be of great importance during LVAD use. When patients supported with LVADs develop RV dysfunction, morbidity and mortality increase dramatically suggesting an important role played by the RV in maintaining forward circulation through the pulmonary circuit and eventually filling the LVAD. In many patients, RV afterload is reduced due to the effective emptying of the left atrium/ventricle during LVAD operation. Right ventricular failure is more likely to develop in patients in whom the pulmonary artery pressure and pulmonary vascular resistance do not decrease during LVAD support [4, 13]. The mechanisms for RV failure during LV assist are still being investigated and likely are multiple: increased venous return leading to RV overload, the loss of the contribution to RV function by the interventricular septum, and changes in pulmonary vascular resistance [9]. One other possible important mechanism is an imbalance of RV blood flow supply and demand. The RV ischemia may lead to RV dysfunction and failure of forward flow through the pulmonary circulation. Two conditions often exist in patients undergoing LVAD implantation both of which might alter supply and demand leading to RV ischemia: (1) presence of RCA atherosclerotic stenoses limiting flow and (2) increased flow demand secondary to an increased wall tension from RV dilatation in the presence of persistently elevated RV afterload. The changes that occur in RV blood flow during LVAD use have not been well documented in the past. It was with the question of the importance of RV blood to RV function that this study was undertaken to provide a baseline for RV blood flow during LVAD use.
Several canine studies suggest that the normal RCA flow is 0.37 to 0.71 mL • min-1 • g-1 with either 36% to 50% of flow occurring during systole [14–17]. However, the dog has a left dominant coronary anatomy compared with right dominant systems generally seen in pigs and humans [18, 19]. The swine model would be a better choice for assessment of right ventricular blood flow characteristics and therefore, was chosen for these experiments.
Under control conditions, total RCA flow is approximately 46 mL/min with 63% of flow occurring in diastole. Generally, aortic pressure is greater than RV pressure, which should allow for continuous RV flow throughout the cardiac cycle. However, the greater flow in diastole demonstrated in these experiments would suggest that systolic myocardial compression may impede coronary flow in the RV similar to but to a lesser extent than in the LV.
During either SYN or ASYN LVAD operation, total RCA flow was similar to the above at approximately 52.6 mL/min but greater flow occurs during the systole of the device. This timing difference is most likely attributable to the fact that systole of the device is almost twice as long as the native cardiac cycle due to the full/fill-full/eject cycle or the 2:1 synchronization of the device. The same flow occurs through the RCA regardless of timing/modality of the device. Regional RV blood flow were similar between control and ASYN modes, but were somewhat higher in SYN modes despite similar arterial and RV pressures.
The PA constriction resulted in increased RV regional flows and increased RV rate–pressure products indicating an increased oxygen supply and demand, although the differences reached statistical significance only in the ASYN mode. The RV flow increased to a maximum of 1.31 mL • min-1 • g-1 in the distal RV region during SYN mode, which is an increase of 41%. The RV flow increased to 1.2 mL • min-1 • g-1 in the proximal RV region representing a 62% increase from control flow. This demonstrates that PA constriction caused a trend to an increase in RV regional flow but more importantly demonstrates a greater capacity of the regions closer to the RCA to augment flow than the distal region, which by definition in this study is farthest away from the artery. There were no changes in RV function as measured by the RVSW and dP/dt during manipulations of RV flow. It is not known from this study whether or not the RV can increase its flows to levels greater than those demonstrated here.
One of the limitations of this study was that it is very difficult to quantify oxygen demand for the right ventricle. Oxygen consumption can be calculated based on arterial and coronary sinus blood oxygen contents but it is difficult to acquire effectively venous blood from the RV. Rate–pressure products have been used to estimate the oxygen demand of the RV [20, 21] but unfortunately this measurement cannot be broken down to its regional components for a comparison to regional blood flows. During PA constriction, stroke volume decreased whereas RV pressure increased. This resulted in an external RV stroke work that was similar to those calculated for the LVAD modes without PA constriction. However, there was likely an increase in internal RV work that was not measured, which may account for the increase in RV blood flow due to increased demand.
Normal pig hearts were used for LVAD support, whereas patients often have coronary artery disease that may affect blood flow to the RV. The gradient of blood flow along the RCA distribution may be of greater importance in patients with RCA stenoses. The most distal areas supplied by the vessels including the RV apex and posterior septum may be affected significantly. In addition, RV conformation is altered with leftward interventricular septal movement during LVAD support. The RV volume and internal diameter are increased leading to greater wall tension and myocardial oxygen consumption as wall tension increases. Therefore, the LVAD patient presents a unique set of variables that may have direct impact on RV blood flow. Further investigation into the pathologic states associated with patients requiring LVADs is needed to assess their impact on RV blood flow and RV mechanical function.
In conclusion, LVADs do not alter regional RV blood flow in normal hearts. There is, however, a gradient of blood flow within the RV that follows the distribution of the RCA. Clinical pathology relating specifically to LVAD patients may result in alterations that potentially may adversely affect RV blood flow leading to RV dysfunction seen in this setting. Further studies with RCA constriction and RV failure are needed to define the importance of RV blood flow to RV function during LVAD support.
|
Acknowledgments
|
|---|
We thank Richard Seymour and Dr James E. Calvin for their assistance. These studies were supported by grants from the Ontario Heart and Stroke Foundation and Bickell Foundation. Drs Hendry and Nathan are supported by Career Scientist Awards from the Ontario Ministry of Health.
|
Footnotes
|
|---|
Address reprint requests to Dr Hendry, University of Ottawa Heart Institute, Rm H207, 1053 Carling Ave, Ottawa, Ontario, Canada K1Y 4E9.
|
References
|
|---|
- Pae WE, Jr. Ventricular assist devices and total artificial hearts: a combined registry experience. Ann Thorac Surg 1993;55:295–8.[Abstract]
- Farrar DJ, Hill JD. Univentricular and biventricular Thoratec VAD support as a bridge to transplantation. Ann Thorac Surg1993;55:276–82.[Abstract]
- Pennington DG, Kanter KR, McBride LR, et al. Seven years' experience with the Pierce-Donachy ventricular assist device. J Thorac Cardiovasc Surg1988;96:901–11.[Abstract]
- Sato N, Mohri H, Miura M, Watanabe T, Nitta S, Sato S. Right ventricular failure during clinical use of a left ventricular assist device. ASAIO Trans1989;35:550–2.[Medline]
- Portner PM, Oyer PE, Pennington DG, et al. Implantable electrical left ventricular assist system: bridge to transplantation and the future. Ann Thorac Surg 1989;47:142–50.[Abstract]
- Ramasamy N, Portner PM. Novacor LVAS: results with bridge to transplant and chronic support. Cardiac Surg 1993;7:363–76.
- Brooks H, Kirk ES, Vokonas PS, Urschel CW, Sonnenblick EH. Performance of the right ventricle under stress: relation to right coronary flow. J Clin Invest 1971;50:2176–83.[Medline]
- Calvin JE, Ascah KJ. Impact of leftward septal shift and potential role of ischemia in its production during experimental right ventricular pressure overload. J Crit Care 1992;7:106–17.
- Farrar DJ, Compton PG, Hershon JJ, Fonger JD, Hill JD. Right heart interaction with the mechanically assisted left heart. World J Surg 1985;9:89–102.[Medline]
- Kawai A, Kormos RL, Mandarino WA, et al. Differential regional function of the right ventricle during the use of a left ventricular assist device. ASAIO Trans 1992;38:M676–8.
- Chow E, Farrar DJ. Effects of left ventricular pressure reductions on right ventricular systolic performance. Am J Physiol 1989;257:H1878–85.[Medline]
- Hendry PJ, Ascah KJ, Rajagopalan K, Calvin JE. Does septal position affect right ventricular function during left ventricular assist in an experimental porcine model. Circulation 1994;90:353–8.
- Kormos RL, Gasior T, Antaki J, et al. Evaluation of right ventricular function during clinical left ventricular assistance. ASAIO Trans 1989;35:547–9.[Medline]
- Lowensohn HS, Khouri EM, Gregg DE, Pyle RL, Patterson RE. Phasic right coronary artery blood flow in conscious dogs with normal and elevated right ventricular pressures. Circ Res 1976;39:760–6.[Abstract/Free Full Text]
- Hess DS, Bache RJ. Transmural right ventricular myocardial blood flow during systole in the awake dog. Circ Res 1979;45:88–94.[Free Full Text]
- Oakley C. Importance of right ventricular function in congestive heart failure. Am J Cardiol 1988;62:14A–19A.[Medline]
- Urabe Y, Tomoike H, Ohzono K, Koyanagi S, Nakamura M. Role of afterload in determining regional right ventricular performance during coronary underperfusion in dogs. Circ Res 1985;57:96–104.[Abstract/Free Full Text]
- Brooks H, Holland R, Al-Sadir J. Right ventricular performance during ischemia: an anatomic and hemodynamic analysis. Am J Physiol 1977;233:H505–13.[Medline]
- Chow E, Foppiano L, Farrar DJ. Regional contractile performance during acute ischemia in porcine right ventricle. Am J Physiol Heart Circ Physiol 1992;263:H135–40.[Abstract/Free Full Text]
- Fixler DE, Monroe GA, Wheeler JM. Hemodynamic alterations during septal or right ventricular ischemia in dogs. Am Heart J 1977;93:210–5.[Medline]
- Reller MD, Morton MJ, Giraud GD, Wu DE, Thornburg KL. Severe right ventricular pressure loading in fetal sheep augments global myocardial blood flow to submaximal levels. Circulation 1992;86:581–8.[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
S. B. Reeder, A. A. Holmes, E. R. McVeigh, and J. R. Forder
Simultaneous Noninvasive Determination of Regional Myocardial Perfusion and Oxygen Content in Rabbits: Toward Direct Measurement of Myocardial Oxygen Consumption at MR Imaging
Radiology,
September 1, 1999;
212(3):
739 - 747.
[Abstract]
[Full Text]
|
|
|
Top
Abstract
Introduction
