- Email: [email protected]
Amrinone versus conventional therapy in pulmonary hypertensive patients awaiting cardiac transplantation
Amrinone Versus Conventional Therapy in Pulmonary Hypertensive Patients Awaiting Cardiac Transplantation G. Michael Deeb, MD, Steven F. Bolling, MD, Todd P. Guynn, MD, and John M. Nicklas, MD The University of Michigan Medical Center, Ann Arbor, Michigan
Pulmonary hypertension is associated with an increased perioperative mortality for orthotopic heart transplantation. A transpulmonary gradient greater than 15 mm Hg or a pulmonary vascular resistance greater than 5 Woods units increases mortality secondary to right heart failure. This study compares amrinone with conventional therapy in 38 transplant candidates with pulmonary hypertension. All patients had elevated transpulmonary gradient, pulmonary vascular resistance, or both. Group 1 (n = 21) received prolonged continuous intravenous amrinone therapy, whereas group 2 (n = 16) received highdose oral diuretics, digitalis, and captopril. Both groups 1 and 2 had decreased pulmonary hypertension, transpulmonary gradient, and pulmonary vascular resistance. However, amrinone was more effective, with a 86% response rate versus 63% response for conventional ther-
apy. Survival awaiting transplantation was significantly higher in group 1 (20of 22,91%) than in group 2 (10of 16, 63%). Although both groups 1 and 2 had significantly decreased pulmonary vascular resistance, only group 2 had significantly decreased systemic vascular resistance. Comparison of pulmonary vascular resistance after therapy showed that the response in group 1 (amrinone) was significantly lower than the response in group 2 (conventional therapy), suggesting that amrinone may function as a direct vasodilator of the pulmonary vasculature. There were no operative deaths or episodes of perioperative right heart failure in either group. Amrinone appears to be more effective and safe than conventional therapy in the treatment of prospective heart transplant candidates with pulmonary hypertension. (Ann Thorac Surg 1989;48:665-9)
P
conventional therapy (digitalis, peripheral vasodilators, and high-dose diuretics) in an outpatient setting versus inpatient continuous intravenous amrinone in 38 pulmonary hypertensive patients awaiting cardiac transplantation.
ulmonary hypertension (PHT) and elevated pulmonary vascular resistance (PVR) increase perioperative mortality for orthotopic heart transplantation. Kormos and associates [l] noted that a transpulmonary gradient (TPG) greater than 15 mm Hg or a PVR greater than 5 Woods units, or both, were associated with an operative mortality of 25% secondary to acute right heart failure. Kirklin and associates [2], using analysis of variance for risk factors in 61 orthotopic heart transplant patients, showed that elevated PVR causes a significant increase in perioperative mortality. Amrinone, a phosphodiesterase inhibitor, was shown by Bolling and colleagues [3] to lower TPG to less than 15 mm Hg and PVR to less than 5 Woods units in 89% of patients when administered with a prolonged intravenous course. The perioperative mortality in this group of patients was 5% (31. Due to the shortage of suitable donor organs and an average wait of approximately 8 months until transplantation, prolonged intravenous therapy in the hospital setting is not feasible, and more conventional modes of therapy with oral medication in an outpatient setting may be more appropriate. This article is a retrospective review comparing use of Accepted for publication July 28, 1989 Presented in part at the Sixtieth Annual Scientific Session of the American Heart Association, Washington, DC, November 14-15, 1988. Address reprint requests to Dr Deeb, Section of Thoracic Surgery, The University of Michigan, 2124F-Taubman Center, Box 0344, Ann Arbor, MI 48109.
0 1989 by The
Society of Thoracic Surgeons
Material and Methods Table 1is a patient profile comparing the demographics of the two groups. There were no significant differences in age, sex, or cause of disease between the two groups. All patients for transplantation were initially evaluated with right heart catheterization. Patients with PHT, defined as either a TPG greater than 15 mm Hg or PVR greater than 5 Woods units, were administered oxygen and given a trial of intravenous nitroprusside therapy. Those demonstrating an adequate decrease in left ventricular alterload reduction with a subsequent decrease in systemic systolic arterial pressure to less than 90 mm Hg but not showing a significant decrease in pulmonary arterial pressure, TPG, and PVR were admitted to the intensive care unit for more intensive treatment. These select patients were believed to have PHT not immediately responsive to left ventricular unloading. Thirty-eight patients who were referred for cardiac transplantation demonstrated PHT independent of immediate left ventricular unloading. Group 1 consisted of 22 patients who were treated as inpatients with prolonged intravenous amrinone therapy. Group 2 consisted of 16 0003-4975/89/$3.50
666
DEEBETAL AMRINONE VERSUS CONVENTIONAL THERAPY
Table 1 . Patient Profile Variable
Grouu 1
U-value
Grouu 2
No. of patients 22 NS 16 51 (59-37) 53 (62-41) Age (yr)" Sex 20 M,2 F 14 M,2 F Cause of disease 14 (67%) 13 (81%) Ischemic Cardiomyopathy 7 (33%) 3 (19%) Responders 19 (86%) 0.053 10 (63%) Days until 3.5 (0.5-7) C0.05 8.2 (2-14) response" 37 (3-136) days 8 (1.5-18)mo Treatment duration (days)' Survival 20 (91%) <0.05 10 (63%) Percentage of 16/19(84%) 7/10(70%) responders transplanted Deaths 2 (10%) Responders 3 (33%) Nonresponders 0 3 (50%) a
Numbers in parentheses are ranges.
patients who received digitalis, high-dose diuretics, and an angiotensin-converting enzyme inhibitor (as an unloading agent) in the hospital for 48 hours and were then discharged and followed as outpatients. All patients had clinically significant congestive heart failure (New York Heart Association class 111to IV) with primary myocardial disease and no intracardiac shunts. All patients were initially admitted to the intensive care unit, and a Swan-Ganz catheter and peripheral arterial line were inserted. Baseline hemodynamic values were obtained, including systemic arterial and venous pressure, pulmonary arterial pressure, pulmonary capillary wedge pressure, cardiac output (CO), transpulmonary gradient, systemic and pulmonary vascular resistances, and daily weights. Measurement of serum electrolytes, blood urea nitrogen, and creatinine, tests of liver function, and platelet counts were performed daily for the first three days and then on a weekly basis. Group 1 patients were administered amrinone in a 0.75-mgikg loading dose and were then started on a continuous infusion of 2.5 pgkglrnin. Hemodynamic variables were measured hourly for four hours and then every four hours for 72 hours. Infusion was titrated to an upper dose of 20 pgikg/min to obtain a desired target effect of a TPG of less than 15 mm Hg, a PVR less than 3 Woods units, or both. After 48 hours the monitoring lines were removed and were reinserted only weekly for 1 month and then monthly if a patient became hemodynamically unstable. All patients were continued on intravenous amrinone until transplantation, hemodynamic decompensation, or exclusion from transplantation for other reasons. If a patient became hemodynamically unstable and required additional inotropic or mechanical support, all subsequent pulmonary pressures were excluded from analysis in this study, but the final outcome of the patient was considered in the statistical assessment.
Ann Thorac Surg 1989:48:665-9
Group 2 patients (n = 16) were also admitted to the intensive care unit and underwent the same monitoring. All received digoxin (0.25 mg/day) and diuretics titrated to weight, levels of blood urea nitrogen and creatinine, and fluid balance. Patients also received an oral angiotensinconverting enzyme inhibitor (captopril) until a systemic blood pressure of approximately 90 mm Hg was achieved or symptoms of decreased peripheral perfusion occurred. Hemodynamic variables were also measured hourly for four hours and then every four hours for 72 hours, at which time the monitoring lines were removed and the patients were discharged and catheterized on a weekly basis for 1 month and then monthly; patients were readmitted if they became hemodynamically unstable. Blood studies were monitored on a weekly basis. Table 2 compares the drug regimens of the two groups. Patients in group 1 received intravenous amrinone at an average dose of 9 pgkglmin. Nine of the patients were taking diuretics, and the mean furosemide dose was 40 mg/day. Two patients received 5 mg/day of metolazone in addition to furosemide. All 16 patients in group 2 received furosemide at a mean dose of 360 mg/day. In addition, 9 patients received metolazone (10 mg/day) and 5 received bumetanide (6 mg/day). All 16 patients in group 2 received digoxin (0.25 mg/day), and 11 received the angiotensin-converting enzyme captopril at an average dose of 37.5 mg/day. The duration of treatment for group 1 ranged from three to 136 days (mean duration, 29 days). Treatment duration for group 2 was 1.5 to 18 months (mean duration, 8 months).
Results Table 1 displays the patients' response to therapy. Nineteen of the 22 patients (86%) in group 1 responded to therapy with a TPG of less than 15 mm Hg or a PVR of less than 3 Woods units, or both, whereas 10 of 16 patients (63%)in group 2 responded (p = 0.053). Group 1 response time was significantly less than that of group 2 (3.5 days versus 8.2 days; p < 0.05). The pretransplantation
Table 2 . Drug Regimens of Patients With Pulmonary Hypertension Group 2 (Conventional Therapy)
Group 1 (Amrinone Therapy)
Dose ~
~~
Lasix Digoxin Captopd Metolazone Bumetanide Amrinone
~~
40 mglday
... ...
5 mglday
No. of Patients
Dose (mg/ day)
No. of Patients
360 0.025 37.5 10 6
16 16 11 9 5
~
9
... ... 2
...
...
0.75 m@g loading, 9 pgkglmin
22
...
...
Ann Thorac Surg 1989:48:665-9
DEEBETAL AMRINONE VERSUS CONVENTIONAL THERAPY
Table 3. Amrinone Versus Conventional Therapy: Pretherapy Hemodynamic Values Hemodynamic Variable BUN (mg/dL) CREAT (mg/dL) P A M (mm Hg) PCWP (mm Hg) TPG (mm Hg) CO (Umin) PVR (WU) MAP (mm Hg) CVP (mm Hg) SVR (WU)
Conventional Therapy (n = 16)
p Value
NS NS
27.7 f 3.5 1.4 f 0.1 46.2 f 1.8 27.7 f 1.8 18.1 f 1.0 3.6 f 0.2 5.1 f 0.4 85 3.6 11.7 f 2.0 21.2 f 1.3
NS NS NS NS NS NS NS NS
*
Amrinone Therapy (n = 22)
28.8 ? 2.2 1.35 f 0.06 44.3 f 1.5 25.2 f 1.0 19.0 f 1.2 3.5 f 0.3 5.8 f 0.3 79.0 f 3.4 12.4 f 2.0 19.0 f 2.2
BUN = blood urea nitrogen; C O = cardiac output; CREAT = creatinine; CVP = central venous pressure; MAP = mean arterial NS = not significant; PAM = mean pulmonary artery pressure; PCWP = pulmonary capillary wedge pressure; PVR = pressure; pulmonary vascular resistance; SVR = systemic vascular resistance; TPG = transpulmonary gradient; WU = Woods units.
survival was also greater in group 1(20 of 22; 91%)than in group 2 (10 of 16; 63%; p < 0.05). There were no perioperative deaths in patients who received transplants in either group. Analysis of all data was performed using paired Student’s t test and analysis of variance. There was no significant difference between the two groups in preoperative level of blood urea nitrogen or creatinine. There was no significant difference in any pretreatment variable between the two groups (Table 3). Table 4 compares hemodynamic variables for group 1 before and after therapy. There was a significant decrease in pulmonary arterial pressure, pulmonary capillary wedge pressure, and PVR, as well as a significant increase in CO. The
Table 4 . Amrinone Therapy (n
=
22)
P
Hemodynamic Variable
Pretherapy
Value
Posttherapy
BUN (mg/dL) CREAT (mg/dL) P A M (mm Hg) PCWP (mm Hg) TPG (mm Hg) CO (Umin) PVR (WU) MAP (mm Hg) CVP (mm Hg) SVR (WU)
28.8 f 2.3 1.3 f 0.06 44.3 f 1.6 25.2 f 1.1 19.0 f 1.2 3.5 f 0.3 5.87 2 0.4 79.0 f 3.4 12.5 f 1.9 20.8 f 2.1
NS NS o.Ooo1
30.2 f 1.6 1.39 ? 0.07 29.0 f 1.5 18.5 f 1.2 10.9 f 0.9 4.56 f 0.3 2.49 f 0.2 80.3 f 3.0 8.3 f 1.8 16.4 f 1.2
0.0001
o.Ooo1 0.015 o.Ooo1 NS NS NS
BUN = blood urea nitrogen; C O = cardiac output; CREAT = creatinine; CVP = central venous pressure; MAP = mean arterial NS = not significant; PAM = mean pulmonary artery pressure; PCWP = pulmonary capillary wedge pressure; PVR = pressure; pulmonary vascular resistance; SVR = systemic vascular resistance; TPG = transpulmonary gradient; WU = Woods units.
667
Table 5 . Conventional Therapy ( n = 26)
P
Hemodynamic Variable
Pretherapy
Value
Posttherapy
BUN (mgldL) CREAT (mgldL) P A M (mm Hg) PCWP (mm Hg) TPG (mm Hg) CO (L/min) PVR (WU) MAP (mm Hg) CVP (mm Hg) SVR (WU) Weight (kg)
27.7 f 3.5 1.4 f 0.1 46.2 f 1.8 27.7 f 1.8 18.2 f 1.0 3.6 f 0.2 5.1 f 0.4 85.0 2 3.6 11.7 f 1.9 21.2 f 1.3 84.6 f 32.6
NS NS 0.0001 0.002 0.003 0.03 0.003 0.03 0.017 0.02 NS
30.2 f 2.5 1.39 f 0.09 32.2 2 2.2 18.9 f 1.8 13.4 f 1.0 4.2 f 0.15 3.3 f 0.3 75.7 f 2.0 5.7 f 1.3 16.6 f 1.1 78.9 f 3.4
BUN = blood urea nitrogen; C O = cardiac output; CREAT = neatinine; CVP = central venous pressure; MAP = mean arterial NS = not significant; PAM = mean pulmonary artery pressure; pressure; PCWP = pulmonary capillary wedge pressure; PVR = pulmonary vascular resistance; SVR = systemic vascular resistance; TPG = transpulmonary gradient; WU = Woods units.
preload (central venous pressure [CVP]) and afterload (systemic vascular resistance) were also reduced, but did not attain statistical significance. Table 5 compares pretreatment and posttreatment results for group 2 (conventional therapy). Pulmonary arterial pressure, pulmonary capillary wedge pressure, and PVR also decreased significantly, and there was a significant increase in CO. The preload (CVP)and afterload (systemicvascular resistance) were significantly reduced in this group. Comparison of the posttreatment variables between the two groups established that the PVR reduction in group 1 was significantly better than that of group 2. In no other measured variable was there a significant difference after therapy. Analysis of the data from both groups was performed comparing those patients who responded to either therapy versus those who did not benefit from therapy (ie, patients who did not have a decrease in TPG, PVR, or PHT during treatment). Between the “responders” and “nonresponders,” the only pretreatment variable of significance was CO. Mean CO in responders was 3.92 L/min, and in nonresponders, it was 3.07 Umin. The survival of all patients from both groups was compared. The only pretreatment variable with a significant difference was CVP. Mean CVP of survivors was 7.7 mm Hg versus 19.3 mm Hg in nonsurvivors. Two of 22 patients (9%) in group 1 had a decrease in platelet count to less than 30,000/mL. Both patients had an initial count greater than 200,000/mL, but it decreased slowly between six and eight days and both patients required transfusion. The decrease was not transient; 1 patient received three separate platelet transfusions and the second patient was transfused at the time of transplantation. Both patients were weaned successfully off amrinone after transplantation, neither had serious postoperative bleeding requiring reoperation, and their plate-
668
DEEBETAL AMRINONE VERSUS CONVENTIONAL THERAPY
let counts returned to more than 200,0001mL within ten days.
Comment The presence of PHT and elevated PVR is a contraindication to orthotopic cardiac transplantation. Elevated PVR causes acute right heart failure in the transplanted heart secondary to an increased right ventricular afterload. Eichhorn and associates [4] showed the effects of phosphodiesterase inhibitors on right ventricular function using simultaneous radionuclide-hemodynamic studies. The phosphodiesterase inhibitor caused a 24% increase in cardiac index without a change in heart rate. The right ventricular preload was unchanged, and the ejection fraction was increased. With the increase in cardiac index and ejection fraction, the pulmonary arterial pressure decreased, indicating a substantial right ventricular afterload reduction. In a previous study, Bolling and associates [3]demonstrated a significant reduction in TPG, PVR, and pulmonary artery mean pressure using prolonged intravenous amrinone therapy in precardiac transplantation candidates. Cardiac output increased significantly, yet there was no significant decrease in systemic arterial pressure or vascular resistance. These data support the concept that amrinone has a direct effect in reducing right ventricular afterload. Our data compare the hemodynamic changes between two groups of differently treated patients with end-stage congestive heart failure and PHT. Both therapies were effective in lowering the pulmonary hemodynamic values to a suitable range for transplantation. Conventional therapy (group 2) had significant effects on the systemic hemodynamics, causing both preload and afterload reduction on the heart. The subsequent pulmonary hemodynamic changes may be secondary to the systemic effects. Group 1 (amrinone) patients did not have significant changes in their systemic hemodynamics, yet manifested pulmonary hemodynamic changes suggesting primary action on the pulmonary vasculature. Tems and colleagues [5] have shown that phosphodiesterase inhibitors have significant systemic unloading effects secondary to peripheral vasodilation. Furthermore, other investigators [6] have shown that the effects of phosphodiesterase inhibitors are more vasodilatory than inotropic, because left ventricular end-systolic pressurediameter relation did not change when cardiac index improved and SVR decreased significantly. The discrepancy between our data concerning SVR and cardiac output and the data of others may reflect peripheral systemic hemodynamic compensation due to prolonged intravenous therapy. The initial systemic effects observed by Miller and co-workers [6] may not be consistent with prolonged therapy, yet favorable pulmonary hemodynamic changes may become evident over time. The pulmonary hemodynamic changes are not reflected by changes in total body water because body weight did not change significantly, although redistribution of body water from the lungs may have occurred. Amrinone-treated
Ann Thorac Surg
1989;48:665-9
patients have a decrease in systemic vascular resistance and a subsequent rise in CO. Although this did not reach statistical significance, it may have contributed partially to the changes in pulmonary hemodynamics in the amrinone group. However, comparison of the systemic and pulmonary hemodynamic variables in both groups after therapy showed that the only variable with statistical significance was PVR. Even though both groups had significant reductions in the PVR, that of the amrinone group decreased more in comparison to that of the conventional group (p < 0.05), which supports a direct vasodilatory effect of amrinone on the pulmonary vasculature. Analysis of the data of all patients in an effort to separate responders from nonresponders showed that pretreatment CO was the only significantly different variable (responders: CO = 3.91 L/min; nonresponders: CO = 3.09 L/min). Cardiac output, therefore, may have some predictive value in separating responders from nonresponders. Similarly, in assessing the differences between patients who survive until transplantation and patients who would not, the pretreatment CVP was significantly different (survivors: mean CVP = 8 mm Hg; nonsurvivors: mean CVP = 19 mm Hg) and may be of predictive value. In conclusion, both conventional and amrinone therapy were effective in lowering pulmonary pressure and PVR to an acceptable range for orthotopic heart transplantation. Preoperative survival was significantly higher in the amrinone-treated group, possibly because these patients were in a hospital setting and were monitored closely. There was also a significant difference between the time interval of treatment and actual transplantation (group 1 = 27 days, group 2 = 8.5 months). Group 1 deaths were most often a result of progressive congestive failure, whereas in group 2 patients, sudden death, possibly due to volume contraction and hypokalemic-induced anythmias, was more common. There were no perioperative deaths in the patients who received transplants in either group. Arminone achieved a more effective rate of response but required prolonged continual intravenous therapy. Pulmonary hypertension redeveloped in some patients once intravenous therapy was discontinued, and such patients therefore required continual in-hospital treatment. Currently the average wait for a donor heart is 8 months, making continuous intravenous amrinone therapy impractical. Because conventional therapy poses a significantly greater risk of pretransplantation death, outpatient oral phosphodiesterase inhibitor therapy may be appropriate. A randomized prospective study comparing conventional therapy to outpatient oral phosphodiesterase therapy is indicated.
References 1. Kormos RL, Thompson M, Hardesty RL, et al. Utility of preoperative right heart catheterization data as a predictor of survival after heart transplantation. J Heart Transplant 1986; 5:391. 2. Kirklin JK, Naftel DC, McGiffin DC, McVay RF, Blackstone
Ann Thorac Surg 1989;48:665-9
EH, Karp RB. Analysis of morbid events and risk factors for death after cardiac transplantation. J Am Coll Cardiol 1988; 11:917-24. 3. Bolling SF, Deeb GM, Crowley DC, Badellino MM, Bove EL. Prolonged amrinone therapy prior to orthotopic cardiac transplantation in patients with pulmonary hypertension. Transplant Proc 1988;2075%6. 4. Eichhom EJ, Konstam MA, Weiland DS, et al. Differential effects of milrinone and dobutamine on right ventricular
DEEBETAL AMRINONE VERSUS CONVENTIONAL THERAPY
669
preload, afterload, and systolic performance in congestive heart failure secondary to ischemic or idiopathic dilated cardiomyopathy. Am J Cardiol 1987;60:1329-33. 5. Terns S, Bourdillon PD, Cheng D, et al. Effects of CI-914 in congestive heart failure due to coronary artery disease or idiopathic cardiomyopathy. Am J Cardiql 1986;58:496400. 6. Miller WP, VanderArk CR, Wiederholt P. Effect of oral milrinone on end-systolic relations in patients with severe congestive heart failure. Am J Cardiol 1987;60:842-6.
BOOKS RECEIVED
T h e Works of William Harvey
Translated by Robert Willis, introduction by Arthur C . Guyton Philadelphia, University of Pennsylvania Press, 1989 624 pp, not illustrated
This volume from the University of Pennsylvania Press includes the very famous Exercitatio Anatomica deMotu Cordis et Sanguinis in Animalibus. It also includes his book, Anatomical Exercises on the Generation of Animals, on which Harvey spent two thirds of his life. There is an excellent introduction to this book by Guyton, and the Robert Willis translations from the original Latin, which were made in 1847 and are considered the most accurate, have been used. This is a book that cardiothoracic surgeons will find
interesting. The introduction by Dr Guyton is as lucid and informative writing as one finds in his books on physiology. Interventional Cardiology Edited by David R. Holmes, jr, atrd Ronald E . Vlietstra Philadelphia, F.A. Davis, 1988 400 pp, illustrated, $70.00 This is a monograph from the Department of Cardiology at the Mayo Clinic that describes their experience, and a short chapter by Dr Hartzell Schaff on the role of the surgeon in interventional cardiology. This clearly gives the viewpoint of the Mayo Clinic.
Pulmonary hypertension is associated with an increased perioperative mortality for orthotopic heart transplantation. A transpulmonary gradient greater than 15 mm Hg or a pulmonary vascular resistance greater than 5 Woods units increases mortality secondary to right heart failure. This study compares amrinone with conventional therapy in 38 transplant candidates with pulmonary hypertension. All patients had elevated transpulmonary gradient, pulmonary vascular resistance, or both. Group 1 (n = 21) received prolonged continuous intravenous amrinone therapy, whereas group 2 (n = 16) received highdose oral diuretics, digitalis, and captopril. Both groups 1 and 2 had decreased pulmonary hypertension, transpulmonary gradient, and pulmonary vascular resistance. However, amrinone was more effective, with a 86% response rate versus 63% response for conventional ther-
apy. Survival awaiting transplantation was significantly higher in group 1 (20of 22,91%) than in group 2 (10of 16, 63%). Although both groups 1 and 2 had significantly decreased pulmonary vascular resistance, only group 2 had significantly decreased systemic vascular resistance. Comparison of pulmonary vascular resistance after therapy showed that the response in group 1 (amrinone) was significantly lower than the response in group 2 (conventional therapy), suggesting that amrinone may function as a direct vasodilator of the pulmonary vasculature. There were no operative deaths or episodes of perioperative right heart failure in either group. Amrinone appears to be more effective and safe than conventional therapy in the treatment of prospective heart transplant candidates with pulmonary hypertension. (Ann Thorac Surg 1989;48:665-9)
P
conventional therapy (digitalis, peripheral vasodilators, and high-dose diuretics) in an outpatient setting versus inpatient continuous intravenous amrinone in 38 pulmonary hypertensive patients awaiting cardiac transplantation.
ulmonary hypertension (PHT) and elevated pulmonary vascular resistance (PVR) increase perioperative mortality for orthotopic heart transplantation. Kormos and associates [l] noted that a transpulmonary gradient (TPG) greater than 15 mm Hg or a PVR greater than 5 Woods units, or both, were associated with an operative mortality of 25% secondary to acute right heart failure. Kirklin and associates [2], using analysis of variance for risk factors in 61 orthotopic heart transplant patients, showed that elevated PVR causes a significant increase in perioperative mortality. Amrinone, a phosphodiesterase inhibitor, was shown by Bolling and colleagues [3] to lower TPG to less than 15 mm Hg and PVR to less than 5 Woods units in 89% of patients when administered with a prolonged intravenous course. The perioperative mortality in this group of patients was 5% (31. Due to the shortage of suitable donor organs and an average wait of approximately 8 months until transplantation, prolonged intravenous therapy in the hospital setting is not feasible, and more conventional modes of therapy with oral medication in an outpatient setting may be more appropriate. This article is a retrospective review comparing use of Accepted for publication July 28, 1989 Presented in part at the Sixtieth Annual Scientific Session of the American Heart Association, Washington, DC, November 14-15, 1988. Address reprint requests to Dr Deeb, Section of Thoracic Surgery, The University of Michigan, 2124F-Taubman Center, Box 0344, Ann Arbor, MI 48109.
0 1989 by The
Society of Thoracic Surgeons
Material and Methods Table 1is a patient profile comparing the demographics of the two groups. There were no significant differences in age, sex, or cause of disease between the two groups. All patients for transplantation were initially evaluated with right heart catheterization. Patients with PHT, defined as either a TPG greater than 15 mm Hg or PVR greater than 5 Woods units, were administered oxygen and given a trial of intravenous nitroprusside therapy. Those demonstrating an adequate decrease in left ventricular alterload reduction with a subsequent decrease in systemic systolic arterial pressure to less than 90 mm Hg but not showing a significant decrease in pulmonary arterial pressure, TPG, and PVR were admitted to the intensive care unit for more intensive treatment. These select patients were believed to have PHT not immediately responsive to left ventricular unloading. Thirty-eight patients who were referred for cardiac transplantation demonstrated PHT independent of immediate left ventricular unloading. Group 1 consisted of 22 patients who were treated as inpatients with prolonged intravenous amrinone therapy. Group 2 consisted of 16 0003-4975/89/$3.50
666
DEEBETAL AMRINONE VERSUS CONVENTIONAL THERAPY
Table 1 . Patient Profile Variable
Grouu 1
U-value
Grouu 2
No. of patients 22 NS 16 51 (59-37) 53 (62-41) Age (yr)" Sex 20 M,2 F 14 M,2 F Cause of disease 14 (67%) 13 (81%) Ischemic Cardiomyopathy 7 (33%) 3 (19%) Responders 19 (86%) 0.053 10 (63%) Days until 3.5 (0.5-7) C0.05 8.2 (2-14) response" 37 (3-136) days 8 (1.5-18)mo Treatment duration (days)' Survival 20 (91%) <0.05 10 (63%) Percentage of 16/19(84%) 7/10(70%) responders transplanted Deaths 2 (10%) Responders 3 (33%) Nonresponders 0 3 (50%) a
Numbers in parentheses are ranges.
patients who received digitalis, high-dose diuretics, and an angiotensin-converting enzyme inhibitor (as an unloading agent) in the hospital for 48 hours and were then discharged and followed as outpatients. All patients had clinically significant congestive heart failure (New York Heart Association class 111to IV) with primary myocardial disease and no intracardiac shunts. All patients were initially admitted to the intensive care unit, and a Swan-Ganz catheter and peripheral arterial line were inserted. Baseline hemodynamic values were obtained, including systemic arterial and venous pressure, pulmonary arterial pressure, pulmonary capillary wedge pressure, cardiac output (CO), transpulmonary gradient, systemic and pulmonary vascular resistances, and daily weights. Measurement of serum electrolytes, blood urea nitrogen, and creatinine, tests of liver function, and platelet counts were performed daily for the first three days and then on a weekly basis. Group 1 patients were administered amrinone in a 0.75-mgikg loading dose and were then started on a continuous infusion of 2.5 pgkglrnin. Hemodynamic variables were measured hourly for four hours and then every four hours for 72 hours. Infusion was titrated to an upper dose of 20 pgikg/min to obtain a desired target effect of a TPG of less than 15 mm Hg, a PVR less than 3 Woods units, or both. After 48 hours the monitoring lines were removed and were reinserted only weekly for 1 month and then monthly if a patient became hemodynamically unstable. All patients were continued on intravenous amrinone until transplantation, hemodynamic decompensation, or exclusion from transplantation for other reasons. If a patient became hemodynamically unstable and required additional inotropic or mechanical support, all subsequent pulmonary pressures were excluded from analysis in this study, but the final outcome of the patient was considered in the statistical assessment.
Ann Thorac Surg 1989:48:665-9
Group 2 patients (n = 16) were also admitted to the intensive care unit and underwent the same monitoring. All received digoxin (0.25 mg/day) and diuretics titrated to weight, levels of blood urea nitrogen and creatinine, and fluid balance. Patients also received an oral angiotensinconverting enzyme inhibitor (captopril) until a systemic blood pressure of approximately 90 mm Hg was achieved or symptoms of decreased peripheral perfusion occurred. Hemodynamic variables were also measured hourly for four hours and then every four hours for 72 hours, at which time the monitoring lines were removed and the patients were discharged and catheterized on a weekly basis for 1 month and then monthly; patients were readmitted if they became hemodynamically unstable. Blood studies were monitored on a weekly basis. Table 2 compares the drug regimens of the two groups. Patients in group 1 received intravenous amrinone at an average dose of 9 pgkglmin. Nine of the patients were taking diuretics, and the mean furosemide dose was 40 mg/day. Two patients received 5 mg/day of metolazone in addition to furosemide. All 16 patients in group 2 received furosemide at a mean dose of 360 mg/day. In addition, 9 patients received metolazone (10 mg/day) and 5 received bumetanide (6 mg/day). All 16 patients in group 2 received digoxin (0.25 mg/day), and 11 received the angiotensin-converting enzyme captopril at an average dose of 37.5 mg/day. The duration of treatment for group 1 ranged from three to 136 days (mean duration, 29 days). Treatment duration for group 2 was 1.5 to 18 months (mean duration, 8 months).
Results Table 1 displays the patients' response to therapy. Nineteen of the 22 patients (86%) in group 1 responded to therapy with a TPG of less than 15 mm Hg or a PVR of less than 3 Woods units, or both, whereas 10 of 16 patients (63%)in group 2 responded (p = 0.053). Group 1 response time was significantly less than that of group 2 (3.5 days versus 8.2 days; p < 0.05). The pretransplantation
Table 2 . Drug Regimens of Patients With Pulmonary Hypertension Group 2 (Conventional Therapy)
Group 1 (Amrinone Therapy)
Dose ~
~~
Lasix Digoxin Captopd Metolazone Bumetanide Amrinone
~~
40 mglday
... ...
5 mglday
No. of Patients
Dose (mg/ day)
No. of Patients
360 0.025 37.5 10 6
16 16 11 9 5
~
9
... ... 2
...
...
0.75 m@g loading, 9 pgkglmin
22
...
...
Ann Thorac Surg 1989:48:665-9
DEEBETAL AMRINONE VERSUS CONVENTIONAL THERAPY
Table 3. Amrinone Versus Conventional Therapy: Pretherapy Hemodynamic Values Hemodynamic Variable BUN (mg/dL) CREAT (mg/dL) P A M (mm Hg) PCWP (mm Hg) TPG (mm Hg) CO (Umin) PVR (WU) MAP (mm Hg) CVP (mm Hg) SVR (WU)
Conventional Therapy (n = 16)
p Value
NS NS
27.7 f 3.5 1.4 f 0.1 46.2 f 1.8 27.7 f 1.8 18.1 f 1.0 3.6 f 0.2 5.1 f 0.4 85 3.6 11.7 f 2.0 21.2 f 1.3
NS NS NS NS NS NS NS NS
*
Amrinone Therapy (n = 22)
28.8 ? 2.2 1.35 f 0.06 44.3 f 1.5 25.2 f 1.0 19.0 f 1.2 3.5 f 0.3 5.8 f 0.3 79.0 f 3.4 12.4 f 2.0 19.0 f 2.2
BUN = blood urea nitrogen; C O = cardiac output; CREAT = creatinine; CVP = central venous pressure; MAP = mean arterial NS = not significant; PAM = mean pulmonary artery pressure; PCWP = pulmonary capillary wedge pressure; PVR = pressure; pulmonary vascular resistance; SVR = systemic vascular resistance; TPG = transpulmonary gradient; WU = Woods units.
survival was also greater in group 1(20 of 22; 91%)than in group 2 (10 of 16; 63%; p < 0.05). There were no perioperative deaths in patients who received transplants in either group. Analysis of all data was performed using paired Student’s t test and analysis of variance. There was no significant difference between the two groups in preoperative level of blood urea nitrogen or creatinine. There was no significant difference in any pretreatment variable between the two groups (Table 3). Table 4 compares hemodynamic variables for group 1 before and after therapy. There was a significant decrease in pulmonary arterial pressure, pulmonary capillary wedge pressure, and PVR, as well as a significant increase in CO. The
Table 4 . Amrinone Therapy (n
=
22)
P
Hemodynamic Variable
Pretherapy
Value
Posttherapy
BUN (mg/dL) CREAT (mg/dL) P A M (mm Hg) PCWP (mm Hg) TPG (mm Hg) CO (Umin) PVR (WU) MAP (mm Hg) CVP (mm Hg) SVR (WU)
28.8 f 2.3 1.3 f 0.06 44.3 f 1.6 25.2 f 1.1 19.0 f 1.2 3.5 f 0.3 5.87 2 0.4 79.0 f 3.4 12.5 f 1.9 20.8 f 2.1
NS NS o.Ooo1
30.2 f 1.6 1.39 ? 0.07 29.0 f 1.5 18.5 f 1.2 10.9 f 0.9 4.56 f 0.3 2.49 f 0.2 80.3 f 3.0 8.3 f 1.8 16.4 f 1.2
0.0001
o.Ooo1 0.015 o.Ooo1 NS NS NS
BUN = blood urea nitrogen; C O = cardiac output; CREAT = creatinine; CVP = central venous pressure; MAP = mean arterial NS = not significant; PAM = mean pulmonary artery pressure; PCWP = pulmonary capillary wedge pressure; PVR = pressure; pulmonary vascular resistance; SVR = systemic vascular resistance; TPG = transpulmonary gradient; WU = Woods units.
667
Table 5 . Conventional Therapy ( n = 26)
P
Hemodynamic Variable
Pretherapy
Value
Posttherapy
BUN (mgldL) CREAT (mgldL) P A M (mm Hg) PCWP (mm Hg) TPG (mm Hg) CO (L/min) PVR (WU) MAP (mm Hg) CVP (mm Hg) SVR (WU) Weight (kg)
27.7 f 3.5 1.4 f 0.1 46.2 f 1.8 27.7 f 1.8 18.2 f 1.0 3.6 f 0.2 5.1 f 0.4 85.0 2 3.6 11.7 f 1.9 21.2 f 1.3 84.6 f 32.6
NS NS 0.0001 0.002 0.003 0.03 0.003 0.03 0.017 0.02 NS
30.2 f 2.5 1.39 f 0.09 32.2 2 2.2 18.9 f 1.8 13.4 f 1.0 4.2 f 0.15 3.3 f 0.3 75.7 f 2.0 5.7 f 1.3 16.6 f 1.1 78.9 f 3.4
BUN = blood urea nitrogen; C O = cardiac output; CREAT = neatinine; CVP = central venous pressure; MAP = mean arterial NS = not significant; PAM = mean pulmonary artery pressure; pressure; PCWP = pulmonary capillary wedge pressure; PVR = pulmonary vascular resistance; SVR = systemic vascular resistance; TPG = transpulmonary gradient; WU = Woods units.
preload (central venous pressure [CVP]) and afterload (systemic vascular resistance) were also reduced, but did not attain statistical significance. Table 5 compares pretreatment and posttreatment results for group 2 (conventional therapy). Pulmonary arterial pressure, pulmonary capillary wedge pressure, and PVR also decreased significantly, and there was a significant increase in CO. The preload (CVP)and afterload (systemicvascular resistance) were significantly reduced in this group. Comparison of the posttreatment variables between the two groups established that the PVR reduction in group 1 was significantly better than that of group 2. In no other measured variable was there a significant difference after therapy. Analysis of the data from both groups was performed comparing those patients who responded to either therapy versus those who did not benefit from therapy (ie, patients who did not have a decrease in TPG, PVR, or PHT during treatment). Between the “responders” and “nonresponders,” the only pretreatment variable of significance was CO. Mean CO in responders was 3.92 L/min, and in nonresponders, it was 3.07 Umin. The survival of all patients from both groups was compared. The only pretreatment variable with a significant difference was CVP. Mean CVP of survivors was 7.7 mm Hg versus 19.3 mm Hg in nonsurvivors. Two of 22 patients (9%) in group 1 had a decrease in platelet count to less than 30,000/mL. Both patients had an initial count greater than 200,000/mL, but it decreased slowly between six and eight days and both patients required transfusion. The decrease was not transient; 1 patient received three separate platelet transfusions and the second patient was transfused at the time of transplantation. Both patients were weaned successfully off amrinone after transplantation, neither had serious postoperative bleeding requiring reoperation, and their plate-
668
DEEBETAL AMRINONE VERSUS CONVENTIONAL THERAPY
let counts returned to more than 200,0001mL within ten days.
Comment The presence of PHT and elevated PVR is a contraindication to orthotopic cardiac transplantation. Elevated PVR causes acute right heart failure in the transplanted heart secondary to an increased right ventricular afterload. Eichhorn and associates [4] showed the effects of phosphodiesterase inhibitors on right ventricular function using simultaneous radionuclide-hemodynamic studies. The phosphodiesterase inhibitor caused a 24% increase in cardiac index without a change in heart rate. The right ventricular preload was unchanged, and the ejection fraction was increased. With the increase in cardiac index and ejection fraction, the pulmonary arterial pressure decreased, indicating a substantial right ventricular afterload reduction. In a previous study, Bolling and associates [3]demonstrated a significant reduction in TPG, PVR, and pulmonary artery mean pressure using prolonged intravenous amrinone therapy in precardiac transplantation candidates. Cardiac output increased significantly, yet there was no significant decrease in systemic arterial pressure or vascular resistance. These data support the concept that amrinone has a direct effect in reducing right ventricular afterload. Our data compare the hemodynamic changes between two groups of differently treated patients with end-stage congestive heart failure and PHT. Both therapies were effective in lowering the pulmonary hemodynamic values to a suitable range for transplantation. Conventional therapy (group 2) had significant effects on the systemic hemodynamics, causing both preload and afterload reduction on the heart. The subsequent pulmonary hemodynamic changes may be secondary to the systemic effects. Group 1 (amrinone) patients did not have significant changes in their systemic hemodynamics, yet manifested pulmonary hemodynamic changes suggesting primary action on the pulmonary vasculature. Tems and colleagues [5] have shown that phosphodiesterase inhibitors have significant systemic unloading effects secondary to peripheral vasodilation. Furthermore, other investigators [6] have shown that the effects of phosphodiesterase inhibitors are more vasodilatory than inotropic, because left ventricular end-systolic pressurediameter relation did not change when cardiac index improved and SVR decreased significantly. The discrepancy between our data concerning SVR and cardiac output and the data of others may reflect peripheral systemic hemodynamic compensation due to prolonged intravenous therapy. The initial systemic effects observed by Miller and co-workers [6] may not be consistent with prolonged therapy, yet favorable pulmonary hemodynamic changes may become evident over time. The pulmonary hemodynamic changes are not reflected by changes in total body water because body weight did not change significantly, although redistribution of body water from the lungs may have occurred. Amrinone-treated
Ann Thorac Surg
1989;48:665-9
patients have a decrease in systemic vascular resistance and a subsequent rise in CO. Although this did not reach statistical significance, it may have contributed partially to the changes in pulmonary hemodynamics in the amrinone group. However, comparison of the systemic and pulmonary hemodynamic variables in both groups after therapy showed that the only variable with statistical significance was PVR. Even though both groups had significant reductions in the PVR, that of the amrinone group decreased more in comparison to that of the conventional group (p < 0.05), which supports a direct vasodilatory effect of amrinone on the pulmonary vasculature. Analysis of the data of all patients in an effort to separate responders from nonresponders showed that pretreatment CO was the only significantly different variable (responders: CO = 3.91 L/min; nonresponders: CO = 3.09 L/min). Cardiac output, therefore, may have some predictive value in separating responders from nonresponders. Similarly, in assessing the differences between patients who survive until transplantation and patients who would not, the pretreatment CVP was significantly different (survivors: mean CVP = 8 mm Hg; nonsurvivors: mean CVP = 19 mm Hg) and may be of predictive value. In conclusion, both conventional and amrinone therapy were effective in lowering pulmonary pressure and PVR to an acceptable range for orthotopic heart transplantation. Preoperative survival was significantly higher in the amrinone-treated group, possibly because these patients were in a hospital setting and were monitored closely. There was also a significant difference between the time interval of treatment and actual transplantation (group 1 = 27 days, group 2 = 8.5 months). Group 1 deaths were most often a result of progressive congestive failure, whereas in group 2 patients, sudden death, possibly due to volume contraction and hypokalemic-induced anythmias, was more common. There were no perioperative deaths in the patients who received transplants in either group. Arminone achieved a more effective rate of response but required prolonged continual intravenous therapy. Pulmonary hypertension redeveloped in some patients once intravenous therapy was discontinued, and such patients therefore required continual in-hospital treatment. Currently the average wait for a donor heart is 8 months, making continuous intravenous amrinone therapy impractical. Because conventional therapy poses a significantly greater risk of pretransplantation death, outpatient oral phosphodiesterase inhibitor therapy may be appropriate. A randomized prospective study comparing conventional therapy to outpatient oral phosphodiesterase therapy is indicated.
References 1. Kormos RL, Thompson M, Hardesty RL, et al. Utility of preoperative right heart catheterization data as a predictor of survival after heart transplantation. J Heart Transplant 1986; 5:391. 2. Kirklin JK, Naftel DC, McGiffin DC, McVay RF, Blackstone
Ann Thorac Surg 1989;48:665-9
EH, Karp RB. Analysis of morbid events and risk factors for death after cardiac transplantation. J Am Coll Cardiol 1988; 11:917-24. 3. Bolling SF, Deeb GM, Crowley DC, Badellino MM, Bove EL. Prolonged amrinone therapy prior to orthotopic cardiac transplantation in patients with pulmonary hypertension. Transplant Proc 1988;2075%6. 4. Eichhom EJ, Konstam MA, Weiland DS, et al. Differential effects of milrinone and dobutamine on right ventricular
DEEBETAL AMRINONE VERSUS CONVENTIONAL THERAPY
669
preload, afterload, and systolic performance in congestive heart failure secondary to ischemic or idiopathic dilated cardiomyopathy. Am J Cardiol 1987;60:1329-33. 5. Terns S, Bourdillon PD, Cheng D, et al. Effects of CI-914 in congestive heart failure due to coronary artery disease or idiopathic cardiomyopathy. Am J Cardiql 1986;58:496400. 6. Miller WP, VanderArk CR, Wiederholt P. Effect of oral milrinone on end-systolic relations in patients with severe congestive heart failure. Am J Cardiol 1987;60:842-6.
BOOKS RECEIVED
T h e Works of William Harvey
Translated by Robert Willis, introduction by Arthur C . Guyton Philadelphia, University of Pennsylvania Press, 1989 624 pp, not illustrated
This volume from the University of Pennsylvania Press includes the very famous Exercitatio Anatomica deMotu Cordis et Sanguinis in Animalibus. It also includes his book, Anatomical Exercises on the Generation of Animals, on which Harvey spent two thirds of his life. There is an excellent introduction to this book by Guyton, and the Robert Willis translations from the original Latin, which were made in 1847 and are considered the most accurate, have been used. This is a book that cardiothoracic surgeons will find
interesting. The introduction by Dr Guyton is as lucid and informative writing as one finds in his books on physiology. Interventional Cardiology Edited by David R. Holmes, jr, atrd Ronald E . Vlietstra Philadelphia, F.A. Davis, 1988 400 pp, illustrated, $70.00 This is a monograph from the Department of Cardiology at the Mayo Clinic that describes their experience, and a short chapter by Dr Hartzell Schaff on the role of the surgeon in interventional cardiology. This clearly gives the viewpoint of the Mayo Clinic.