- Email: [email protected]
Right ventricular function during left heart bypass
J THoRAc
CARDIOVASC SURG
85:49-53, 1983
Right ventricular function during left heart bypass Right heart failure may occur during mechanical support of the left ventricle (LV). Right ventricular (RV) junctional changes were studied in eight dogs (26.9 ± 1.4 kg) subjected to various degrees of left heart bypass (LHBP) with a roller pump. The venous return to the right atrium was controlled with a second pump. RV function was evaluated by peak RV developed pressure, its first derivative (dp/dt), and mean RA pressure measurements. Left atrial, LV, and aortic pressure and both roller pump flows were determined. Incremental increases in LHBP flow ratio ([LHBP flow X 100] divided by venous return flow) to 60%, 90%, and 100% were associated with decrements in RV dp/dt from the control of 212 ± 17 torr/sec to 192 ± 16, 178 ± 16, and 168 ± 13 torr/sec, respectively. Biventricular bypass or total cardiopulmonary bypass with extracorporeal membrane oxygenation seems to be indicated if LHBP flow ratio greater than 90% is required to maintain adequate body perfusion. Maximal LV decompression to obtain the greatest reduction of LV myocardial oxygen consumption may not be the ideal goal during LV mechanical assistance.
Alfonso Tadaomi Miyamoto, M.D. (by invitation), Shigeo Tanaka, M.D. (by invitation), and Jack M. Matloff, M.D., Los Angeles, Calif.
Right ventricular (R V) failure has been recognized as a complication or a natural sequela of mechanical assistance of the left ventricle (LV).! Intensive inotropic or supplemental mechanical support of the R V is often required as well, either in the form of biventricular bypass" or total cardiopulmonary bypass with extracorporeal membrane oxygenation." The functional interaction between the ventricles has been evaluated by many investigators using varying experimental models. Moulopoulos and associates" described the deleterious effects of R V bypass and R V distension on the performance of the L V in in situ canine hearts. Mantini and colleagues" concluded that partial heart bypass reduces the work of the ipsilateral ventricle but increases volume work of the opposite ventricle. The dependence of the diastolic pressurevolume relationship of one ventricle on the degree of filling of the other has been studied by various researchers with somewhat differing results. In an iso-
lated cat heart preparation in which the R V and LV ejected and filled independently, Elzinga and associates" found that when the atrial filling pressure increased on one side, the output from the opposite side of the heart decreased. Santamore's group," using isolated rabbit hearts beating isovolumically, found that increasing LV volume increased R V diastolic and developed pressures. The purpose of this study was to characterize the R V functional changes occurring during highly effective mechanical assistance of the LV. An ejecting, normal, in situ canine heart model with intact autonomic system innervation was used. Material and methods
Read at the Sixty-second Annual Meeting of The American Association for Thoracic Surgery, Phoenix, Ariz., May 3-5, 1982. Address for reprints: Alfonso T. Miyamoto, M.D., Department of Thoracic and Cardiovascular Surgery, Cedars-Sinai Medical Center, 8700 Beverly Blvd., Los Angeles, Calif. 90048.
Dogs (26.9 ± 1.4 kg) were anesthetized with sodium pentobarbital (25 mg/kg intravenously followed by a constant infusion at 0.1 to 0.15 mg/kg/min) and ventilated with a respirator (Harvard Apparatus Co., Inc., S. Natic, Mass.) to obtain an arterial P0 2 greater than 100 torr and a Pco, of 35 to 45 torr. The heart was exposed via a median sternotomy, and appropriate cannulations and connections were made to allow total cardiopulmonary bypass (Fig. 1) so that a cannula could be sutured in the interatrial septum via a right atriotomy, as previously described." 9 Fig. 1 also depicts the connection in parallel of an additional trans-
© 1983 The C. V. Mosby Co.
49
From the Department of Thoracic and Cardiovascular Surgery, Cedars-Sinai Medical Center, Los Angeles, California.
0022-5223/83/010049+05$00.50/0
The Journal of Thoracic and Cardiovascular
50 Miyamoto. Tanaka. Matloff
Surgery
SEPTAL
CANNULA'----... ---II-----LAP
RAP-H-+--
---il--AOP LVP
RVP -++--+--+--
dp/dt
......1<:"""::"......_-TaLV
CANNULA
Instruments, Inc., Los Angeles, Calif.). Pressures and flows were recorded with a multichannel recorder (Electronics for Medicine, Inc., White Plains, N. Y.). The R V function was evaluated by determining RV contractility as measured by RV dp/dt in the presence of a constant systemic venous return to the right atrium or a preload of approximately 100 ml/kg/min. The LHBP flow ratio ([LHBP flow x 100]/venous retum flow) was varied in stepwise fashion from nil or control, fully loaded LV to maximal LHBP to obtain total LV decompression. A 5 minute period of stabilization was allowed at each of the selected LHBP flow ratios of 60%, 80%, and 90% and at maximal flow rates obtainable by the septal cannula. The effect of maximal LHBP flows obtained by the transapical cannula, as well as total L V decompression obtained by the simultaneous use of the septal and transapical cannulas, was also studied.
Results
LHBP
PUMP
Fig. 1. Experimental model. Systemic blood flow was controlled by the venous return pump to the right atrium, after interruption of both venaecavae. A commonelectromagnetic flow probe was used to measure flows of both pumps sequentially, by proper clamping of the bypass line. RAP, Right atrial pressure. RVP, Right ventricular pressure. LAP, Left atrial pressure. AOP, Aortic pressure. LVP, Left ventricular pressure. TaLV, Transapical left ventricular cannula. LHBP, Left heart bypass. e.O., Cardiac output. apical LV cannula to the inlet tubing of the left heart bypass (LHBP) pump for total LV decompression. The venous return to the right atrium was controlled by diverting both caval flows to the oxygenator and was returned at a constant rate to the right atrium with a second roller pump (Sams, Inc., Ann Arbor, Mich.). Appropriate fluid-filled catheters were placed in the left atrium, LV, aorta, and right atrium and connected to pressure transducers (Bell & Howell Co., Pasadena, Calif.). A tip transducer catheter (Millar Instruments, Inc., Houston, Texas) was placed within the RV chamber through the free wall of the RV for measurement of the developed R V pressure; its rate of pressure development (RV dp/dt) was obtained with a differentiator and recorded. Flows of both pumps, i.e., LHBP pump and venous return pump, were measured with a common electromagnetic flow probe (Micron
The entire series of determinations was completed in eight dogs, and the results are summarized in Table I. As the LHBP flow ratio increased, the mean left atrial pressure decreased and the peak developed LV pressure decreased progressively and drastically to become almost zero at maximal LV decompression. The right atrial pressure decreased initially from a control of 3.3 ± 0.9 torr under fully pre loaded conditions of the LV to 2.3 ±0.5 torr (p > 0.05) at an LHBP flow ratio of 80%. The mean right atrial pressure tended to increase thereafter with further increase of the LHBP flow ratio, to become significantly higher than the initial control pressures (5.1 ± 1.1 torr [p < 0.05]) when total decompression of the LV was achieved. Although the R V peak developed pressure did not change significantly, there was a tendency for it to decrease with increasing LHBP flow ratios. Most strikingly and significantly, the R V dp/dt decreased progressively as the LHBP flow ratios were increased. Of interest is the relatively unchanged R V dp/dt at maximal LHBP ratio when transapical LV cannulation alone was used, even at flows equal to or greater than those obtained at maximal rates by septal cannulation alone. However, there was a clear trend for the right atrial pressure during the transapical mode of LHBP to be greater than that during the septal cannulation mode of LHBP. These findings might be indicative of RV dysfunction during LHBP.
Discussion The LV, being the dominant ventricle, has captured the attention of many investigators. LV function has
Volume 65 Number 1 January, 1963
Right ventricular function during left heart bypass
5I
Table I. Right ventricular function during left heart bypass (LHBP) in eight dogs (mean ± SEM) (LHBPF x 100) VRF (%)
Control LHBP- septal LHBP-septal LHBP- septal LHBP-septal max. LHBP-TaLV max. LHBP-septal + TaLV max.
No LHBP 60 80 90
101 102 105
LAP (torr)
2.9 0.7 0.7 0.3 0.1 1.5 0.1
± ± ± ± ± ± ±
1.2 0.5* 0.4* 0.1* 0.1* 0.7* 0.1*
Peak LVP (torr)
126 109 101 99 90 72 2
± ± ± ± ± ± ±
4.5 5.7* 6.3*t 6.3*t 6.4*t II.O*t 1.7*t
RV dp/dt (torr/sec)
212 192 182 178 174 200 168
± ± ± ± ± ± ±
17 16* 17* 16*t 16*t 15 13*t
RAP (torr)
3.3 2.3 2.3 2.2 2.6 3.4 5.1
± ± ± ± ± ± ±
0.9 0.6 0.5 0.5 0.6 0.8 I.I*t
Peak RVP (torr)
27 26 26 27 26 26 25
± ± ± ± ± ± ±
2.9 3.1 3.0 3.3 3.4 3.0 3.0
Legend: The controlled venous return flow (VRF) to the right atrium was considered as systemic blood flow disregarding the coronary sinus flow. Left heart bypass flow (lHBPF) ratios greater than venous return flow can be explained by the coronary sinus flow and the bronchial flow returning to the left heart. lAP, Left atrial pressure. lVP, Left ventricular pressure. RAP, Right atrial pressure. RvP, Right ventricular pressure. TalV max., Maximal flows obtained by transapical cannulation. Septal + Tal V max., Maximal flows obtained by the combination of septal and transapical cannulation. Statistical analysis: unpaired Student's t test.
*p < 0.05 versus control. tp < 0.05 versus 60% lHBPF.
been the subject of more extensive studies than RV function. Moulopoulos and associates" described the LV performance during various degrees of filling of the RV. During RV bypass, the LVend-diastolic pressure was higher than it would have been at the same LV stroke work but without right heart bypass. LV function curves were displaced to the right, and the rate of development of isometric tension of the LV was reduced during right heart bypass. The extent of these changes was similar in both sympathectomized or nonsympathectomized groups of animals. Interestingly, similar LV functional changes occurred during RV distension beyond the control volumes. The effects of changes in right heart volume on L V function were independent of LV preload or afterload, heart rate, and the presence or absence of sympathetic reflex activity. These authors concluded that the LV seems to work more efficiently when the RV maintains its normal volume. They hypothesized the possible existence of a direct mechanical relationship of the two ventricles, mediated perhaps by the activity of the muscular bundles encircling both ventricles (deep sinospiral and superficial bulbospiral muscles). The dependence of the diastolic pressure-volume relationship of one ventricle on the degree of filling of the other, and especially the effects of RV distension on the LV pressure-volume characteristics, have been studied extensively by a number of investigators under various experimental circumstances, using passively filled noncontracting ventricles'" as well as contracting ventricles.v 7, 10, 11 Likewise, LV volume changes produce significant alterations in R V function. Elzinga and associates, 6 using an isolated cat heart preparation to break the inseries arrangement between the RV and LV, observed
that increasing the filling of the left side of the heart has a significant depressive effect on the right side, and that this depressive effect was diminished by removal of the pericardium. These findings are the mirror image of those described by Moulopoulos and co-workers" of the effects of right heart distension on LV function. Elzinga's group" concluded that an increase in L V diastolic volume is associated with a decrease in RV volume and thus changes the RVend-diastolic pressure-volume relationship. On the other hand, Santamore and associates? used an isovolumically beating isolated rabbit heart preparation and reported that progressive increases of L V volume produced increasing RV diastolic as well as RV isovolumic developed pressures. More recently, Elzinga, Piene, and de Jong;" using an isolated cat heart preparation, described a "crosstalk phenomenon" between the left and right hearts during systole. The right pump function was analyzed by the relationship between the mean RV pressure and the mean RV output obtained during a sudden synchronized and drastic fall in aortic pressure induced during a single heartbeat. This allowed the L V developed pressure to be changed during one single beat. The authors concluded that the RV pump function is dependent on the LV contraction mode; i.e., an isovolumic beat on the left side of the heart enhances RV pump function. The right heart not only is capable of generating more pressure at a given output level, but it also can eject more blood at a certain pressure level in the presence of an isovolumic beat on the left heart. In the present study, the heart was kept in situ, and the left heart ejected whatever volume the LHBP did not remove from the left heart. The LV developed pressure and, concomitantly, its isovolumic pressure were controlled by changing the LHBP flow ratio (or LV
The Journal of
52
Miyamoto, Tanaka, Matloff
preload) rather than by manipulating the arterial pressure or afterload, which was maintained constant. The RV function was analyzed by measuring global contractility changes as determined by the rate of pressure development in the presence of a constant systemic venous return (or preload) to the right heart. Although the pulmonary artery pressures were not measured, the RV developed pressures should correlate closely to the peak pulmonary artery pressures. The RV peak developed pressures remained essentially unchanged, and therefore it could be assumed that the afterload to the R V remained unchanged as well. The initial decrease of the rate of pressure development of the RV observed at an LHBP flow ratio of 60% to 80% was accompanied by a fall of the mean right atrial pressure. This can probably be explained on the basis that the decreased RV contractility (decreased RV dp/dt) is accompanied by increased RV distensibility due to the decreased LV diastolic volume.!" The increased R V distensibility reduces RVend-diastolic pressure and therefore decreases mean right atrial pressure even in the presence of impaired RV contractility. However, at maximal left heart decompression, i.e., when LV pressure is reduced to almost zero, the impairment of the rate of pressure development of the RV reaches a maximum to the point of also significantly increasing the mean right atrial pressure. This increased right atrial pressure occurs even in the presence of a slightly reduced right heart preload, which is assumed to occur on the basis of a 25% decrease in coronary flow induced by total LV decompression, as previously reported. 9 These results confirm those reported by Santamore and associates," in which moderate reduction of LV volume produced decreased RVend-diastolic pressures, and increasing LV volumes increased the RV developed pressures. Our results give further support to the conclusions reached by Elzinga, Piene, and de Jong ," who found that the isovolumic contraction mode of the L V enhanced the pump function of the RV. Whether it is the maximal developed pressure of the isovolumic LV contraction, or the absolute volume, and consequently the geometry of the LV which prevails at the end of diastole or during early systole (isovolumic phase) is a difficult point to be differentiated. To elucidate why the RV dp/dt is relatively unaffected, even at high LHBP ratios, when the transapical cannula alone is used, despite the lower LV developed pressure when compared to that obtained by septal cannulation, needs further work. This suggests that L V geometry might be different whether the left heart output is diverted from the atrium or from the LV itself, and that L V geometry might be the determinant of these effects.
Thoracic and Cardiovascular Surgery
In addition to this direct myocardial negative effect of the LHBP on the RV function, the volume workload increase to the contralateral ventricle in the presence of partial heart bypass must be considered." The combination of both factors might explain the clinical observations by many investigators in this field of overt right-sided failure during mechanical support of the left heart. 1 Often, this right-sided failure develops late during the course of left heart support and is seemingly unrelated to the adequacy of intraoperative myocardial protection. In the light of these findings, the aim of temporary L V support, commonly regarded as maximal reduction of LV myocardial oxygen consumption by maximal decompression of the LV, may require reevaluation. Global cardiac functional recovery is the ultimate goal of cardiac support. If LHBP flow ratio requirements exceed 80% to 90% of the cardiac output, it might be advisable to implement cannulations for additional right heart bypass," i.e., biventricular bypass," or at least to make available the potential for total cardiopulmonary bypass using a membrane oxygenator as an integral part of the support system." An objective of establishing a minimal LHBP flow ratio just sufficient to meet the general body metabolic and perfusion requirements, as advocated early by Bregman,' Litwak," and their associates and reconfirmed more recently by Rose and colleagues;" might be better than the establishment of maximal left heart decompression to reach maximal reduction of myocardial oxygen consumption. REFERENCES
2
3
4
5
Bregman D, Norman JC, Golding L, Litwak R, Penningron G, Peters 1, Shumakov Y, Taguchi K, Turino M: Mechanical cardiac assist/left ventricular cannulation techniques. Panel conference. Trans Am Soc Artif Intern Organs 25:503-509, 1979 Taguchi K, Murashital, Nakagaki M, Mochizuki T, Matsumura M, Isono M, Hamanaka Y, Tsuchiya T: Clinical application of biventricular bypass with six consecutive patients. Trans Am Soc Artif Intern Organs 26:428-43 I, 1980 Bartlett RH, Gazzaniga AB, Wetmore NE, Rucker R, Huxtable RF: Extracorporeal membrane oxygenation (ECMO) in the treatment of cardiac and respiratory failure in children. Trans Am Soc Artif Intern Organs 26:578-579, 1980 Moulopoulos SD, Sarcas A, Stamatelopoulos S, Arealis E: Left ventricular performance during bypass or distension of the right ventricle. Circ Res 17:484-491, 1965 Mantini E, Tanaka S, Horta-Da-Silva P, Lillehei CW:
Volume 85 Number 1 January, 1983
6
7
8
9
10
II
12
13
14
15
Some aspects of altered physiology during partial right and left ventricular bypass. Trans Am Soc Artif Intern Organs 8:288-291, 1967 Elzinga G, Van Grondelle R, Westerhof N, van den Bos GC: Ventricular interference. Am 1 PhysioI226:941-947, 1974 Santamore WP, Lynch PR, Meier G, Heckmans 1, Bove AA: Myocardial interaction between the ventricles. 1 Appl Physiol 41:362-368, 1976 Miyamoto AT, Tanaka S, Robinson LF, Matloff 1M: A new cannulation technique for atrio-aortic left heart support bypass. Atrial septum. Trans Am Soc Artif Intern Organs 26:466-469, 1980 Miyamoto AT, Tanaka S, Matloff 1M: Myocardial O 2 consumption (M V0 2 ) during left heart bypass by atrial septal suture cannulation. Trans Am Soc Artif Intern Organs 27:495-498, 1981 Taylor RR, Covell lW, Sonnenblick EH, Ross 1 Jr: Dependence of ventricular distensibility on filling of the opposite ventricle. Am 1 Physiol 213:711-718, 1967 Bemis CE, Serur lR, Borkenhagen D, Sonnenblik EH, Urschel CW: Influence of right ventricular filling pressure on left ventricular pressure and dimension. Circ Res 34:498-504, 1974 Elzinga G, Piene H, de long IP: Left and right ventricular pump function and consequences of having two pumps in one heart. A study on the isolated cat heart. Circ Res 46:564-574, 1980 Miyamoto AT, Tanaka S, Matloff 1M: A new atrial cannulation for total biventricular bypass. Am Soc Artif Intern Organs Abstracts 11:7, 1982 Litwak R, Koffsky R, lurado R, Lukban SB, Ortiz AF, Fischer AP, Sherman 11, Silvay G, Lajam FA: Use of a left heart assist device after intra-cardiac surgery. Technique and clinical experience. Ann Thorac Surg 21: 191202, 1976 Rose DM, Colvin SB, Culliford AT, Cunningham IN, Adams PX, Glassman E, Isom OW, Spencer FC: Longterm survival with partial left heart bypass following perioperative myocardial infarction and shock. 1 THoRAc CARDIOVASC SURG 83:483-492, 1982
Right ventricular function during left heart bypass
53
Discussion DR. WILLIAM S. PIERCE Hershey, Pa.
Dr. Miyamoto and his associates have directed our attention to an important problem-the relationship of ventricular interdependency. Extensive experience with the use of L V-to-aortic bypass in animals with normal hearts had failed to identify any significant problem with R V function. However, every group now employing L V assistance in postcardiotomy patients is acutely aware of the importance of an adequately functioning R V. In our series of 22 patients who have had LV assist pumping after cardiac operations, R V failure has been present in six instances. Half of these patients had blood removed from the LV apex while, in the other half, blood was removed from the left atrium. There was no difference in the incidence of R V failure in these two groups. We agree with the authors that there are minor derangements in RV function associated with LV decompression. However, our observations suggest that R V failure is primarily a result of R V muscle dysfunction, secondary to poor myocardial preservation or ischemia of the RV myocardium. I would ask Dr. Miyamoto to comment on the mechanism whereby LV decompression depresses R V function. Second, do you believe that the minor derangements in R V function that you have demonstrated have clinical importance? DR. M I YAM 0 T 0 (Closing) Dr. Pierce, I appreciate your comments. I believe the mechanism for R V functional impairment is the septal shift that occurs with L V volume changes. We are currently trying to identify this shift with two-dimensional echocardiography. Indeed, the more complete the LHBP decompression is, the greater the septal shift is and the worse the R V function becomes. In regard to clinical implications, I believe that, in most patients, the R V functional impairment could be treated with inotropic agents. However, if the need for LV bypass flow ratio is greater than 90% of cardiac output, it probably is best to have capabilities for R V bypass implemented at the time the LHBP is placed. In that way, the need to add a second bypass after the chest has been closed could be obviated.
CARDIOVASC SURG
85:49-53, 1983
Right ventricular function during left heart bypass Right heart failure may occur during mechanical support of the left ventricle (LV). Right ventricular (RV) junctional changes were studied in eight dogs (26.9 ± 1.4 kg) subjected to various degrees of left heart bypass (LHBP) with a roller pump. The venous return to the right atrium was controlled with a second pump. RV function was evaluated by peak RV developed pressure, its first derivative (dp/dt), and mean RA pressure measurements. Left atrial, LV, and aortic pressure and both roller pump flows were determined. Incremental increases in LHBP flow ratio ([LHBP flow X 100] divided by venous return flow) to 60%, 90%, and 100% were associated with decrements in RV dp/dt from the control of 212 ± 17 torr/sec to 192 ± 16, 178 ± 16, and 168 ± 13 torr/sec, respectively. Biventricular bypass or total cardiopulmonary bypass with extracorporeal membrane oxygenation seems to be indicated if LHBP flow ratio greater than 90% is required to maintain adequate body perfusion. Maximal LV decompression to obtain the greatest reduction of LV myocardial oxygen consumption may not be the ideal goal during LV mechanical assistance.
Alfonso Tadaomi Miyamoto, M.D. (by invitation), Shigeo Tanaka, M.D. (by invitation), and Jack M. Matloff, M.D., Los Angeles, Calif.
Right ventricular (R V) failure has been recognized as a complication or a natural sequela of mechanical assistance of the left ventricle (LV).! Intensive inotropic or supplemental mechanical support of the R V is often required as well, either in the form of biventricular bypass" or total cardiopulmonary bypass with extracorporeal membrane oxygenation." The functional interaction between the ventricles has been evaluated by many investigators using varying experimental models. Moulopoulos and associates" described the deleterious effects of R V bypass and R V distension on the performance of the L V in in situ canine hearts. Mantini and colleagues" concluded that partial heart bypass reduces the work of the ipsilateral ventricle but increases volume work of the opposite ventricle. The dependence of the diastolic pressurevolume relationship of one ventricle on the degree of filling of the other has been studied by various researchers with somewhat differing results. In an iso-
lated cat heart preparation in which the R V and LV ejected and filled independently, Elzinga and associates" found that when the atrial filling pressure increased on one side, the output from the opposite side of the heart decreased. Santamore's group," using isolated rabbit hearts beating isovolumically, found that increasing LV volume increased R V diastolic and developed pressures. The purpose of this study was to characterize the R V functional changes occurring during highly effective mechanical assistance of the LV. An ejecting, normal, in situ canine heart model with intact autonomic system innervation was used. Material and methods
Read at the Sixty-second Annual Meeting of The American Association for Thoracic Surgery, Phoenix, Ariz., May 3-5, 1982. Address for reprints: Alfonso T. Miyamoto, M.D., Department of Thoracic and Cardiovascular Surgery, Cedars-Sinai Medical Center, 8700 Beverly Blvd., Los Angeles, Calif. 90048.
Dogs (26.9 ± 1.4 kg) were anesthetized with sodium pentobarbital (25 mg/kg intravenously followed by a constant infusion at 0.1 to 0.15 mg/kg/min) and ventilated with a respirator (Harvard Apparatus Co., Inc., S. Natic, Mass.) to obtain an arterial P0 2 greater than 100 torr and a Pco, of 35 to 45 torr. The heart was exposed via a median sternotomy, and appropriate cannulations and connections were made to allow total cardiopulmonary bypass (Fig. 1) so that a cannula could be sutured in the interatrial septum via a right atriotomy, as previously described." 9 Fig. 1 also depicts the connection in parallel of an additional trans-
© 1983 The C. V. Mosby Co.
49
From the Department of Thoracic and Cardiovascular Surgery, Cedars-Sinai Medical Center, Los Angeles, California.
0022-5223/83/010049+05$00.50/0
The Journal of Thoracic and Cardiovascular
50 Miyamoto. Tanaka. Matloff
Surgery
SEPTAL
CANNULA'----... ---II-----LAP
RAP-H-+--
---il--AOP LVP
RVP -++--+--+--
dp/dt
......1<:"""::"......_-TaLV
CANNULA
Instruments, Inc., Los Angeles, Calif.). Pressures and flows were recorded with a multichannel recorder (Electronics for Medicine, Inc., White Plains, N. Y.). The R V function was evaluated by determining RV contractility as measured by RV dp/dt in the presence of a constant systemic venous return to the right atrium or a preload of approximately 100 ml/kg/min. The LHBP flow ratio ([LHBP flow x 100]/venous retum flow) was varied in stepwise fashion from nil or control, fully loaded LV to maximal LHBP to obtain total LV decompression. A 5 minute period of stabilization was allowed at each of the selected LHBP flow ratios of 60%, 80%, and 90% and at maximal flow rates obtainable by the septal cannula. The effect of maximal LHBP flows obtained by the transapical cannula, as well as total L V decompression obtained by the simultaneous use of the septal and transapical cannulas, was also studied.
Results
LHBP
PUMP
Fig. 1. Experimental model. Systemic blood flow was controlled by the venous return pump to the right atrium, after interruption of both venaecavae. A commonelectromagnetic flow probe was used to measure flows of both pumps sequentially, by proper clamping of the bypass line. RAP, Right atrial pressure. RVP, Right ventricular pressure. LAP, Left atrial pressure. AOP, Aortic pressure. LVP, Left ventricular pressure. TaLV, Transapical left ventricular cannula. LHBP, Left heart bypass. e.O., Cardiac output. apical LV cannula to the inlet tubing of the left heart bypass (LHBP) pump for total LV decompression. The venous return to the right atrium was controlled by diverting both caval flows to the oxygenator and was returned at a constant rate to the right atrium with a second roller pump (Sams, Inc., Ann Arbor, Mich.). Appropriate fluid-filled catheters were placed in the left atrium, LV, aorta, and right atrium and connected to pressure transducers (Bell & Howell Co., Pasadena, Calif.). A tip transducer catheter (Millar Instruments, Inc., Houston, Texas) was placed within the RV chamber through the free wall of the RV for measurement of the developed R V pressure; its rate of pressure development (RV dp/dt) was obtained with a differentiator and recorded. Flows of both pumps, i.e., LHBP pump and venous return pump, were measured with a common electromagnetic flow probe (Micron
The entire series of determinations was completed in eight dogs, and the results are summarized in Table I. As the LHBP flow ratio increased, the mean left atrial pressure decreased and the peak developed LV pressure decreased progressively and drastically to become almost zero at maximal LV decompression. The right atrial pressure decreased initially from a control of 3.3 ± 0.9 torr under fully pre loaded conditions of the LV to 2.3 ±0.5 torr (p > 0.05) at an LHBP flow ratio of 80%. The mean right atrial pressure tended to increase thereafter with further increase of the LHBP flow ratio, to become significantly higher than the initial control pressures (5.1 ± 1.1 torr [p < 0.05]) when total decompression of the LV was achieved. Although the R V peak developed pressure did not change significantly, there was a tendency for it to decrease with increasing LHBP flow ratios. Most strikingly and significantly, the R V dp/dt decreased progressively as the LHBP flow ratios were increased. Of interest is the relatively unchanged R V dp/dt at maximal LHBP ratio when transapical LV cannulation alone was used, even at flows equal to or greater than those obtained at maximal rates by septal cannulation alone. However, there was a clear trend for the right atrial pressure during the transapical mode of LHBP to be greater than that during the septal cannulation mode of LHBP. These findings might be indicative of RV dysfunction during LHBP.
Discussion The LV, being the dominant ventricle, has captured the attention of many investigators. LV function has
Volume 65 Number 1 January, 1963
Right ventricular function during left heart bypass
5I
Table I. Right ventricular function during left heart bypass (LHBP) in eight dogs (mean ± SEM) (LHBPF x 100) VRF (%)
Control LHBP- septal LHBP-septal LHBP- septal LHBP-septal max. LHBP-TaLV max. LHBP-septal + TaLV max.
No LHBP 60 80 90
101 102 105
LAP (torr)
2.9 0.7 0.7 0.3 0.1 1.5 0.1
± ± ± ± ± ± ±
1.2 0.5* 0.4* 0.1* 0.1* 0.7* 0.1*
Peak LVP (torr)
126 109 101 99 90 72 2
± ± ± ± ± ± ±
4.5 5.7* 6.3*t 6.3*t 6.4*t II.O*t 1.7*t
RV dp/dt (torr/sec)
212 192 182 178 174 200 168
± ± ± ± ± ± ±
17 16* 17* 16*t 16*t 15 13*t
RAP (torr)
3.3 2.3 2.3 2.2 2.6 3.4 5.1
± ± ± ± ± ± ±
0.9 0.6 0.5 0.5 0.6 0.8 I.I*t
Peak RVP (torr)
27 26 26 27 26 26 25
± ± ± ± ± ± ±
2.9 3.1 3.0 3.3 3.4 3.0 3.0
Legend: The controlled venous return flow (VRF) to the right atrium was considered as systemic blood flow disregarding the coronary sinus flow. Left heart bypass flow (lHBPF) ratios greater than venous return flow can be explained by the coronary sinus flow and the bronchial flow returning to the left heart. lAP, Left atrial pressure. lVP, Left ventricular pressure. RAP, Right atrial pressure. RvP, Right ventricular pressure. TalV max., Maximal flows obtained by transapical cannulation. Septal + Tal V max., Maximal flows obtained by the combination of septal and transapical cannulation. Statistical analysis: unpaired Student's t test.
*p < 0.05 versus control. tp < 0.05 versus 60% lHBPF.
been the subject of more extensive studies than RV function. Moulopoulos and associates" described the LV performance during various degrees of filling of the RV. During RV bypass, the LVend-diastolic pressure was higher than it would have been at the same LV stroke work but without right heart bypass. LV function curves were displaced to the right, and the rate of development of isometric tension of the LV was reduced during right heart bypass. The extent of these changes was similar in both sympathectomized or nonsympathectomized groups of animals. Interestingly, similar LV functional changes occurred during RV distension beyond the control volumes. The effects of changes in right heart volume on L V function were independent of LV preload or afterload, heart rate, and the presence or absence of sympathetic reflex activity. These authors concluded that the LV seems to work more efficiently when the RV maintains its normal volume. They hypothesized the possible existence of a direct mechanical relationship of the two ventricles, mediated perhaps by the activity of the muscular bundles encircling both ventricles (deep sinospiral and superficial bulbospiral muscles). The dependence of the diastolic pressure-volume relationship of one ventricle on the degree of filling of the other, and especially the effects of RV distension on the LV pressure-volume characteristics, have been studied extensively by a number of investigators under various experimental circumstances, using passively filled noncontracting ventricles'" as well as contracting ventricles.v 7, 10, 11 Likewise, LV volume changes produce significant alterations in R V function. Elzinga and associates, 6 using an isolated cat heart preparation to break the inseries arrangement between the RV and LV, observed
that increasing the filling of the left side of the heart has a significant depressive effect on the right side, and that this depressive effect was diminished by removal of the pericardium. These findings are the mirror image of those described by Moulopoulos and co-workers" of the effects of right heart distension on LV function. Elzinga's group" concluded that an increase in L V diastolic volume is associated with a decrease in RV volume and thus changes the RVend-diastolic pressure-volume relationship. On the other hand, Santamore and associates? used an isovolumically beating isolated rabbit heart preparation and reported that progressive increases of L V volume produced increasing RV diastolic as well as RV isovolumic developed pressures. More recently, Elzinga, Piene, and de Jong;" using an isolated cat heart preparation, described a "crosstalk phenomenon" between the left and right hearts during systole. The right pump function was analyzed by the relationship between the mean RV pressure and the mean RV output obtained during a sudden synchronized and drastic fall in aortic pressure induced during a single heartbeat. This allowed the L V developed pressure to be changed during one single beat. The authors concluded that the RV pump function is dependent on the LV contraction mode; i.e., an isovolumic beat on the left side of the heart enhances RV pump function. The right heart not only is capable of generating more pressure at a given output level, but it also can eject more blood at a certain pressure level in the presence of an isovolumic beat on the left heart. In the present study, the heart was kept in situ, and the left heart ejected whatever volume the LHBP did not remove from the left heart. The LV developed pressure and, concomitantly, its isovolumic pressure were controlled by changing the LHBP flow ratio (or LV
The Journal of
52
Miyamoto, Tanaka, Matloff
preload) rather than by manipulating the arterial pressure or afterload, which was maintained constant. The RV function was analyzed by measuring global contractility changes as determined by the rate of pressure development in the presence of a constant systemic venous return (or preload) to the right heart. Although the pulmonary artery pressures were not measured, the RV developed pressures should correlate closely to the peak pulmonary artery pressures. The RV peak developed pressures remained essentially unchanged, and therefore it could be assumed that the afterload to the R V remained unchanged as well. The initial decrease of the rate of pressure development of the RV observed at an LHBP flow ratio of 60% to 80% was accompanied by a fall of the mean right atrial pressure. This can probably be explained on the basis that the decreased RV contractility (decreased RV dp/dt) is accompanied by increased RV distensibility due to the decreased LV diastolic volume.!" The increased R V distensibility reduces RVend-diastolic pressure and therefore decreases mean right atrial pressure even in the presence of impaired RV contractility. However, at maximal left heart decompression, i.e., when LV pressure is reduced to almost zero, the impairment of the rate of pressure development of the RV reaches a maximum to the point of also significantly increasing the mean right atrial pressure. This increased right atrial pressure occurs even in the presence of a slightly reduced right heart preload, which is assumed to occur on the basis of a 25% decrease in coronary flow induced by total LV decompression, as previously reported. 9 These results confirm those reported by Santamore and associates," in which moderate reduction of LV volume produced decreased RVend-diastolic pressures, and increasing LV volumes increased the RV developed pressures. Our results give further support to the conclusions reached by Elzinga, Piene, and de Jong ," who found that the isovolumic contraction mode of the L V enhanced the pump function of the RV. Whether it is the maximal developed pressure of the isovolumic LV contraction, or the absolute volume, and consequently the geometry of the LV which prevails at the end of diastole or during early systole (isovolumic phase) is a difficult point to be differentiated. To elucidate why the RV dp/dt is relatively unaffected, even at high LHBP ratios, when the transapical cannula alone is used, despite the lower LV developed pressure when compared to that obtained by septal cannulation, needs further work. This suggests that L V geometry might be different whether the left heart output is diverted from the atrium or from the LV itself, and that L V geometry might be the determinant of these effects.
Thoracic and Cardiovascular Surgery
In addition to this direct myocardial negative effect of the LHBP on the RV function, the volume workload increase to the contralateral ventricle in the presence of partial heart bypass must be considered." The combination of both factors might explain the clinical observations by many investigators in this field of overt right-sided failure during mechanical support of the left heart. 1 Often, this right-sided failure develops late during the course of left heart support and is seemingly unrelated to the adequacy of intraoperative myocardial protection. In the light of these findings, the aim of temporary L V support, commonly regarded as maximal reduction of LV myocardial oxygen consumption by maximal decompression of the LV, may require reevaluation. Global cardiac functional recovery is the ultimate goal of cardiac support. If LHBP flow ratio requirements exceed 80% to 90% of the cardiac output, it might be advisable to implement cannulations for additional right heart bypass," i.e., biventricular bypass," or at least to make available the potential for total cardiopulmonary bypass using a membrane oxygenator as an integral part of the support system." An objective of establishing a minimal LHBP flow ratio just sufficient to meet the general body metabolic and perfusion requirements, as advocated early by Bregman,' Litwak," and their associates and reconfirmed more recently by Rose and colleagues;" might be better than the establishment of maximal left heart decompression to reach maximal reduction of myocardial oxygen consumption. REFERENCES
2
3
4
5
Bregman D, Norman JC, Golding L, Litwak R, Penningron G, Peters 1, Shumakov Y, Taguchi K, Turino M: Mechanical cardiac assist/left ventricular cannulation techniques. Panel conference. Trans Am Soc Artif Intern Organs 25:503-509, 1979 Taguchi K, Murashital, Nakagaki M, Mochizuki T, Matsumura M, Isono M, Hamanaka Y, Tsuchiya T: Clinical application of biventricular bypass with six consecutive patients. Trans Am Soc Artif Intern Organs 26:428-43 I, 1980 Bartlett RH, Gazzaniga AB, Wetmore NE, Rucker R, Huxtable RF: Extracorporeal membrane oxygenation (ECMO) in the treatment of cardiac and respiratory failure in children. Trans Am Soc Artif Intern Organs 26:578-579, 1980 Moulopoulos SD, Sarcas A, Stamatelopoulos S, Arealis E: Left ventricular performance during bypass or distension of the right ventricle. Circ Res 17:484-491, 1965 Mantini E, Tanaka S, Horta-Da-Silva P, Lillehei CW:
Volume 85 Number 1 January, 1983
6
7
8
9
10
II
12
13
14
15
Some aspects of altered physiology during partial right and left ventricular bypass. Trans Am Soc Artif Intern Organs 8:288-291, 1967 Elzinga G, Van Grondelle R, Westerhof N, van den Bos GC: Ventricular interference. Am 1 PhysioI226:941-947, 1974 Santamore WP, Lynch PR, Meier G, Heckmans 1, Bove AA: Myocardial interaction between the ventricles. 1 Appl Physiol 41:362-368, 1976 Miyamoto AT, Tanaka S, Robinson LF, Matloff 1M: A new cannulation technique for atrio-aortic left heart support bypass. Atrial septum. Trans Am Soc Artif Intern Organs 26:466-469, 1980 Miyamoto AT, Tanaka S, Matloff 1M: Myocardial O 2 consumption (M V0 2 ) during left heart bypass by atrial septal suture cannulation. Trans Am Soc Artif Intern Organs 27:495-498, 1981 Taylor RR, Covell lW, Sonnenblick EH, Ross 1 Jr: Dependence of ventricular distensibility on filling of the opposite ventricle. Am 1 Physiol 213:711-718, 1967 Bemis CE, Serur lR, Borkenhagen D, Sonnenblik EH, Urschel CW: Influence of right ventricular filling pressure on left ventricular pressure and dimension. Circ Res 34:498-504, 1974 Elzinga G, Piene H, de long IP: Left and right ventricular pump function and consequences of having two pumps in one heart. A study on the isolated cat heart. Circ Res 46:564-574, 1980 Miyamoto AT, Tanaka S, Matloff 1M: A new atrial cannulation for total biventricular bypass. Am Soc Artif Intern Organs Abstracts 11:7, 1982 Litwak R, Koffsky R, lurado R, Lukban SB, Ortiz AF, Fischer AP, Sherman 11, Silvay G, Lajam FA: Use of a left heart assist device after intra-cardiac surgery. Technique and clinical experience. Ann Thorac Surg 21: 191202, 1976 Rose DM, Colvin SB, Culliford AT, Cunningham IN, Adams PX, Glassman E, Isom OW, Spencer FC: Longterm survival with partial left heart bypass following perioperative myocardial infarction and shock. 1 THoRAc CARDIOVASC SURG 83:483-492, 1982
Right ventricular function during left heart bypass
53
Discussion DR. WILLIAM S. PIERCE Hershey, Pa.
Dr. Miyamoto and his associates have directed our attention to an important problem-the relationship of ventricular interdependency. Extensive experience with the use of L V-to-aortic bypass in animals with normal hearts had failed to identify any significant problem with R V function. However, every group now employing L V assistance in postcardiotomy patients is acutely aware of the importance of an adequately functioning R V. In our series of 22 patients who have had LV assist pumping after cardiac operations, R V failure has been present in six instances. Half of these patients had blood removed from the LV apex while, in the other half, blood was removed from the left atrium. There was no difference in the incidence of R V failure in these two groups. We agree with the authors that there are minor derangements in RV function associated with LV decompression. However, our observations suggest that R V failure is primarily a result of R V muscle dysfunction, secondary to poor myocardial preservation or ischemia of the RV myocardium. I would ask Dr. Miyamoto to comment on the mechanism whereby LV decompression depresses R V function. Second, do you believe that the minor derangements in R V function that you have demonstrated have clinical importance? DR. M I YAM 0 T 0 (Closing) Dr. Pierce, I appreciate your comments. I believe the mechanism for R V functional impairment is the septal shift that occurs with L V volume changes. We are currently trying to identify this shift with two-dimensional echocardiography. Indeed, the more complete the LHBP decompression is, the greater the septal shift is and the worse the R V function becomes. In regard to clinical implications, I believe that, in most patients, the R V functional impairment could be treated with inotropic agents. However, if the need for LV bypass flow ratio is greater than 90% of cardiac output, it probably is best to have capabilities for R V bypass implemented at the time the LHBP is placed. In that way, the need to add a second bypass after the chest has been closed could be obviated.