Effects of variable dose milrinone in patients with low cardiac output after cardiac surgery We studied 99 adult patients after elective cardiac surgery who had low cardiac output (cardiac index 8 mm Hg). Patients received milrinone by loading dose (50 wg/kg over a IO-minute period), followed by a continuous infusion of either 0.375, 0.5, or 0.75 wg/kg/min (low-, middle-, and high-dose groups, respectively) given for a minimum of 12 hours. Patients were allocated to each dosage group sequentially, not randomly. Hemodynamic measurements were made before the loading dose and at 15, 30, 45, and 60 minutes, 3, 6, and 12 hours after the start of milrinone therapy. Further measurements were made at 2 and 4 hours after treatment was stopped. Milrinone therapy was associated with a rapid, well-sustained, and highly significant increase in cardiac index in all three dose groups (p < 0.001) and a similar fall occurred in pulmonary capillary wedge pressure in all groups (p < 0.001). Significant increases occurred in heart rate in all three groups (p < 0.001). Systemic and pulmonary vascular resistance also fell significantly, although changes in this latter parameter were less predictable and more dose dependent. Few serious treatment-related adverse effects were seen. We conclude that intravenous milrinone is an effective and safe therapy for the treatment of low output states after cardiac surgery. (Au HEART J 1991;121:1995-9.)
Robert 0. Feneck, for The European London,
United
Multicenter
Trial
Kingdom.
Milrinone is a bipyridine methyl carbonitrile analog of amrinone that has been reported to be 10 to 30 times more potent than amrin0ne.l It is a nonglycosidic, noncatecholamine agent with both positive inotropic and vasodilator effects that have been demonstrated in both experimental animals2 and humans.3 The mechanism of action would appear to be by selective and potent inhibition of the cyclic adenosine monophosphate (CAMP)-specific cardiac phosphodiesterase F III. This leads to an increase in cyclic adenosine monophosphate levels, thereby increasing intracellular calcium availability.4 Milrinone has previously been found to be an effective treatment in patients with chronic heart failure, where it has produced a significant fall in pulmonary capillary wedge pressure (PCWP) and increases in cardiac index (C!I).l, 5, 6 There is no data on the use of milrinone in patients recovering from cardiac surgery, but because the proven effects of phosphodiesterase inhibitors have been shown to include enhanced contractility, a reduction in left venReprint requests: Robert Chest Hospital, Banner *See
Appendix
4/O/28163
Group*
0. Feneck, Consultant Anesthesiologist. Road, London E2, United Kingdom.
fbr participating
investigators
and
centers.
London
tricular afterload, and improvement in left ventricular diastolic compliance ,4 there are good theoretic grounds for the assumption that milrinone may be a useful agent in the treatment of low output states in patients recovering from cardiac surgery. We have therefore investigated its use in such patients. PATIENTS
AND
METHODS
Adult patients scheduled to undergo elective cardiac surgery at one of the five participating centers and who gave written informed consent were eligible to enter the study. The trial had the approval of all the local ethical committees concerned and also the approval of the European Ethical Review Committee. All patients were free of significant primary renal or hepatic disease. No patient had a serious hematologic abnormality or an abnormal platelet count. No long-acting cardiovascular medication was given on the morning of surgery, but patients received their routine sedative anesthetic premeditation, and general anesthesia was established in accordance with routine practice in each of the participating centers. Similarly, cardiopulmonary bypass perfusion and surgical techniques were carried out in accordance with local routine practice and protocols. After surgery, all patients were transferred to the intensive care unit. At this time the patients were monitored by continuous display of the ECG and invasive systemic blood pressure. In addition, right atria1 pressure (RAP), mean 1995
1996
Table
Feneck for The European I. Patient
BSA, graft;
+ MVR
Body surface AVR, atria1
Trial
Group
data
No. of patients Age (yr) M/F BSA (m2) Surgery CABG AVR MVR AVR
Multicenter
Low dose
Mid dose
High dose
34 64 26/8 1.86
34 63 20114 1.79
31 62 20/l 1 1.78
27 3 2 2
20 4 9 1
21 5 3 2
area; M/F, male/female; valve replacement; MVR,
CABG, mitral
coronary artery bypass valve replacement.
pulmonary artery pressure (PAP), and PCWP measurements were made by means of a balloon-tipped, flowdirected, pulmonary artery catheter. Cardiac output was measured by the bolus thermodilution technique in accordance with standard practice, and intermittent blood gas analysis was also carried out. All vascular pressures were measured by suitably calibrated, commercially available transducers and displayed on appropriate bedside monitors. In each patient cardiac output measurements were made at a comparable point during the respiratory cycle. Multiple measurements were made at each time and the mean calculated. After surgery all patients had a 12-lead ECG, a chest radiograph, and a platelet count before they received milrinone. Patients were eligible to receive milrinone if they fulfilled the following criteria: (1) stable with regard to ventilation and gas exchange, sedation and analgesia, and fluid and diuretic requirements; (2) low CI, which was defined as a CI of less than 2.51 L/min/m2; (3) adequate cardiac filling pressure, defined as a minimum PCWP of 8 mm Hg but variable for each patient and determined by each investigative team; and (4) hemodynamic stability, which required two consecutive readings of CI and PCWP to be made within a lo-minute period. These measurements had to be such that the lower of the readings was within 15% of the upper reading. In addition, these measurements had to be made within 60 minutes of receiving treatment with milrinone. At these times heart rate (HR), mean systemic arterial pressure (MAP), PAP, and RAP measurements were recorded, in addition to CI and PCWP. Patients were excluded from the study if they demonstrated any evidence of uncontrolled supraventricular tachyarrhythmia, hemodynamically or prognostically significant ventricular extrasystoles, if they had received any long-acting cardiovascular medication, including vasodilators, or if they demonstrated any untoward response to anesthesia or surgery. Milrinone was administered as an initial loading dose of 50 pg/kg in each patient over a lo-minute period. This was followed by a continuous infusion of either 0.375 rglkglmin (low dose), 0.5 wg/kg/min (mid dose), or 0.75 pg/kg/min
American
June 1991 Heart Journal
(high dose). Patients were sequentially allocated to each dosage group in each center. Thus the low-dose group was completed first, then the mid-dose group, and finally the high-dose group. Further measurements of the hemodynamic parameters described were made after 15,30,45, and 60 minutes and after 3, 6, and 12 hours of milrinone therapy. In addition, measurements were made at 2 and 4 hours after milrinone therapy was stopped. Adverse events were recorded in each of the five centers. Data were recorded to establish whether the adverse event was minor or serious, the intensity of the event (i.e., mild, moderate, or severe), the cause of the event (i.e., drug related or not), and what action if any was needed. Statistical analysis of the data was carried out by one-way analysis of variance and paired and unpaired Student t tests as appropriate. RESULTS
Patient data are shown in Table I. As can be seen, there were no significant differences in age, body surface area or weight between the three groups. There were more women in the middle-dose than in the low-dose group, and the middle-dose group also contained a greater number of patients undergoing mitral valve surgery. Baseline hemodynamic data are shown in Table II. There were no significant differences in baseline values for HR, CI, stroke index (SI), PCWP, systemic vascular resistance (SVR), MAP, RAP, PAP, or pulmonary vascular resistance (PVR), and thus the groups were hemodynamically comparable before receiving the drug. Hemodynamic data during and after milrinone treatment are shown in Tables III to VI; the data are expressed as the mean percentage change in each hemodynamic parameter from its baseline value in each of the three groups. Tests of significance are also shown as within-group tests only. As can be seen, after the loading dose there was a rapid and highly significant increase in CI that was well maintained throughout the treatment period in each of the three groups. This increase in CI persisted even after milrinone was discontinued nificantly greater than baseline values
and was sigin each of the
three groups 4 hours after milrinone therapy had been discontinued (Table III). There was a rapid fall in PCWP in each of the three groups
that
was
highly
significant
for
the first
60
minutes of treatment and remained variably significant
in each
of the three
groups
thereafter.
After
treatment, was stopped, PCWP increased in each group. Thus 4 hours after treatment was stopped, PCWP was higher than baseline in the low-dose group, which was similar to baseline in the middle-
Volume
121
Number
6.
Table
Part
Postcardiac
2
II. Baseline hemodynamic data (mean k SEM)
Hg) 11.6 (0.5) 11.9 (0.5) 11.2 (0.5)
15 Min CI (L/m2) LOW Mid High
81 (2.4) 81 (2.3) 83 (3.3)
PCWP Low Mid
SI (ml/m? Low Mid High
23 (1.0) 24 (1.0) 25 (1.5)
High
86 (2.6) 85 (2.5) 81 (2.9)
+40* (4.2) +30* (4.5) +36* (4.9) Hg) -30* (4.7) -34* (4.5) -35* (4.0)
12
hr
Post
+42* (4.9) +34* (4.5) +44* (4.7)
+58* (6.8) +49* (5.1) +66* (6.5)
+44* (6.3) +27* (3.8) +47* (6.6)O
-20*
-15t (7.2) -2o* (4.3) -15t (6.0)
+201 (6.3) -3 (6.7)
(4.7) -25* (4.1) -22* (4.3)
4.X (6.3)
set
$p < 0.01.
Table
IV. Hemodynamic data % Change (mean r SEM)
cm--5) 1876 (102) 1840 (90) 1698 (84) SI (ml) Low
Hg) 8.4 (0.6) 9.0 (0.5) 9.1 (0.5)
Mid
set
cmm5) 202 (12) 224 (19) 184 (16)
dose group, and lower than baseline in the high-dose group. The increase in CI occurred both as a result of an increase in HR and SI (Table IV). These parameters were significantly increased throughout the treatment period and remained variably but significantly increased thereafter. The effects of treatment on MAP were more variable. As can be seen, MAP fell during treatment. However, the effects of milrinone on SVR were more striking; there was a rapid, wellmaintained, and highly significant reduction in SVR in each of the three groups, and this persisted during and after treatment (Table V). Table VI shows the results of hemodynamic measurements from the right side of the heart. As can be seen, there was a fall in RAP immediately after treatment with the loading dose of milrinone in each
15 Min
60 Min
+22.8* (3.5) +14.6*
+25.6* (4.7) +19.1* (4.0) +25.1* (4.3)
+41.8* (5.7) +35.6* (5.5) +42.2* (4.7)
+32.9* (5.2) +20.7* (4.5) +33.5* (6.6)
+13.4* (1.8) +13.6* (2.8) +15.0*
+11.1* (2.9) +ll.Si (3.3) +17.3* (3.6)
+8.7t (3.6) +7.4t (3.2) +11.41: (3.8)
(4.2) High
PVR (dynes Low Mid High
(mm
60 Min
*p < 0.001 tp < 0.05.
Hg)
SVR (dynes Low Mid High RAP (mm Low Mid High
Ill. Hemodynamic data
Table
1.86 (0.06) 1.95 (0.05) 1.98 (0.06)
(beats/min) Low Mid High
MAP (mm Low Mid High
1997
90 Change (mean k SEM)
CI (L/m21 I,OW Mid High PCWP (mm Low Mid High HR
surgery
+19.5* (4.8)
HR (beats/min) Low +14.4* (1.4) Mid +13.9* (1.9) High +14.5* (2.3)
(2.2)
12 hr
Post
*p < 0.001. tp < 0.05. $p < 0.01.
of the three groups. However, this fall was not maintained during the treatment period, and the changes were not significant. Similarly, the effects of treatment on PAP suggest an initial reduction that is variable both within and between the groups and of limited significance. The effects of treatment on PVR were more reproducible, however, in that there was a highly significant reduction in PVR in each of the three groups after 12 hours of milrinone therapy. Table VII shows the data regarding adverse events during this study. As can be seen, there was no difference between the groups regarding the incidence, nature, or severity of the adverse events seen.
1998
Table
Feneck for The European
V.
Hemodynamic
15 Min
60 Min
MAP (mm Low
(2.2) High
-33.3t (2.5)
Trial Group
data 9; Change
Hg) -5.4* (2.5) Mid -11.7t (2.0) High -13.71 (2.2) SVR (dynes set Low -29.5t (2.4) Mid -29.31
Multicenter
Table (mean
-3.9
-8.6’
(2.3)
(3.8)
+2.3 (3.7) -2.6 (3.4) -2.9 (4.9)
-12.6t
(2.3)
cmm5) -27.3t (3.0) -30.1t
(2.5) -35.1t (2.5)
-39.1t (2.9) -39.1t (2.8) -43.2t (2.8)
Hemodynamic
-25.67 (3.8) -22.0t (4.4) -27.3t (4.9)
‘p < 0.05. tp < o.nn1 $p < 0.01.
DISCUSSION
Previous studies have evaluated the effects of milrinone in patients suffering from chronic heart failure.1*5*6 These data provide a valuable insight into the safety and efficacy of milrinone, and in this study we have used the loading dose and maintenance infusion regimens found to be effective in patients with chronic heart failure. Nevertheless, it is clear that there may be substantial differences between patients who have chronic heart failure and those recovering from cardiac surgery. In the latter group, poor left ventricular function may be attributed to poor preoperative left ventricular function and inadequate myocardial protection, thereby leading to a degree of ischemic or reperfusion injury, or a combination of the above factors linked with the depressant effects of anesthesia and cardiopulmonary bypass. Of particular importance may be the effects of hypothermic cardiopulmonary bypass on diastolic function and ventricular compliance that may not have fully recovered during the early postoperative period. In spite of these differences, treatment with milrinone was associated with a rapid and sustained improvement in CI and PCWP in all three patient groups. These data are in agreement with data from patients with chronic heart failure in whom similar effects were also seen. The increase in CI seen in our study was associated with an increase in SI and HR, but this latter increase, although highly significant, was noted more for the statistical reproducibility of the effect than for its magnitude. Other data also suggest that the chro-
15 Min RAP (mm Low
June 1991 Heart Journal
data % Change
Post
-12.6* (3.1)
VI.
t SEM)
12 hr
-9.11 (1.9) -9.4* (3.1)
American
(mean
-+ SEMI
60 Min
12 hr
PUSi
+13.1 (34.4) -9.6 (5.6) -8.4 (6.9)
+46.3 (56.1) -8.7 (4.1) +51.7 (50.5)
+54.4 (42.8) +18.6 (18.1) +72.8 (63.7)
- 10.3t (3.3) -8.7 (3.6) -14.3’ (3.8)
-9.8 (5.7) -14.2t (3.7) -5.4 (4.5)
+8.1
-24.2* (6.9) -5.3 (11.4) -8.1 (7.3)
-36.01 (5.4) -25.7* (9.4 -26.7t (6.9)
Hg)
Mid
-14.8 (11.7) -26.6
High
-22.1
(4.6) (6.5) MAP (mm Low
Hg) -15.6* (2.8) Mid -17.0* (2.6) High -14.7* (3.2) PVR (dynes set Low -25.6* (3.9) Mid -10.1 (6.0) High -7.9 (6.7)
(4.8) +1.1 (5.2) +11.4 (5.6)
cmm5) -28.71
(7.2) -2.3 (13.9) +2.4 (9.3)
*p < 0.01. tp < 0.05. tp < 0.001
Table
VII.
Adverse
Low, 16/34 (47%) 6 Arrhythmias 5 Nausea *Total of seven arrhythmias nonsustained ventricular tivity.
events* Mid, 16/34 (47%) 5 Arrhythmias 4 Nausea deemed tachycardia
drug related; or multifocal
High, 14131 (45%) 6 Arrhythmias 4 Nausea no evidence ventricular
of sustained/ ectopic ac-
notropic effects of milrinone are mild both in absolute terms and in comparison with other treatments.5 We believe that this is an important point, since tachycardia is well known to have the potential for inducing myocardial oxygen imbalance and has also been shown to be the hemodynamic abnormality that correlates most significantly with the development of perioperative ischemic episodes in surgical patients.l Clearly, from our data milrinone is seen to be an effective systemic vasodilator, and indeed this effect could account substantially for the hemodynamic effects seen in this group of patients. However, the effects of the drug on MAP were not marked, and after 12 hours of milrinone therapy, the mean reductions in MAP were between 8 % and 13% in the three groups. These changes were similar in magnitude to those produced by the loading dose only, although
“Ohw Number
121 6, Part
Postcardiac
2
it should be said that during the loading dose period, transient increases in MAP were also noted, which were quite marked in a very small number of patients. As shown above, the effects of milrinone on RAP, PAP, and PVR were not so marked as the other hemodynamic effects seen, although there were highly significant reductions in PVR in all three groups by the end of the treatment period. Clearly the effectiveness of a pulmonary vasodilator drug is substantially affected by the resting tone of the pulmonary vasculature, and while there were pulmonary hypertensive patients in each of the three treatment groups, the baseline data of the three groups did not suggest that pulmonary hypertension is a feature of the groups as a whole. Thus we will need to separate out the pulmonary hypertensive patients and analyze these patients as a further subgroup before we can reach meaningful conclusions about the effects of milrinone on pulmonary hypertensive surgical patients. From the data in Table VII it can be seen that the incidence of adverse effects was similar in the three groups of patients. However, it should be remembered that the incidence of minor adverse effects in patients recovering from cardiac surgery is also common. In particular, the two adverse effects highlighted in Table VII, arrhythmia and nausea, are particularly common in patients recovering from cardiac surgery. It should be emphasized that these were trivial arrhythmias without hemodynamic effects or prognostic significance and that only five of 16 were possibly drug related. Two adverse events were deemed to be serious by the investigators. These were both episodes of atria1 fibrillation, in one patient occurring de novo, and in the other patient occurring as an acceleration of the ventricular rate in prexisting atria1 fibrillation. In both of these cases milrinone treatment was continued for some hours, and the patients’ outcome was not adversely affected. Both episodes of atria1 fibrillation occurred within 5 minutes of administering the loading dose of milrinone in otherwise stable patients and therefore were deemed to be treatment related. In conclusion, our analysis of this data suggests that the effects of milrinone that we have seen in patients with low cardiac output recovering from elective cardiac surgery are very similar to those seen previously in patients who have chronic heart failure. These beneficial effects were well maintained throughout the treatment period, and the incidence of adverse events was acceptably low.
surgery
1999
REFERENCES
JL, Bairn DS. Fein SA. Goldstein RA, LeJemtel TH, Likoff MJ. Efficacy and safety of sustained (48 hr) intravenous
1. Anderson infusions
of milrinone
failure; a multicenter
in patients
with
severe
congestive
heart
study. J Am Co11Cardiol1987;9:711-22.
2. Alousi AA, Stankus GP, Stuart JC, et al. Characterization the cardiotonic effects of milrinone, a new potent cardiac
of
bipyridine, on isolated tissues from several animal species. J Cardiovasc
3.
Pharmacol
1983;5:804.
Ludmer PL, Wright RF, Arnold MO, et al. Separation of the direct myocardial and vasodilator actions of milrinone administration by an intracoronary 1986;73:13&
infusion
technique.
Circulation
Chatterjee K. Newer oral inotropic agents; phosphodiesterase inhibitors. Crit Care Med 1990;18:S34-8. 5. Biddle TL, Benotti JR, Creager MA, et al. Comparison of intravenous milrinone and dibutamine for congestive heart failure secondary to either ischemic or dilated cardiomyopathy. Am J Cardiol 1987;59:1345-50. 4.
6. Benotti
JR, Lawrence
JL, McCue
inetics and pharmacodynamics gestive 7.
heart
failure.
JE, Alpert
of milrinone
Am J Cardiol
JS. Pharmacok-
in chronic con-
1985;56:685-9.
Slogoff S, Keats AS. Does perioperative myocardial ischemia lead to postoperative myocardial infarction? Anesthesiology 1985;62:107-14.
APPENDIX. THE EUROPEAN TRIAL GROUP
MILRINONE
London Chest Hospital, London, United R. 0. Feneck, Consultant Anesthesiologist A. Hill, Research Fellow, Anesthesia R. K. Walesby, Consultant Cardiothoracic
MULTICENTER Kingdom
Surgeon
National Heart Hospital, London, United Kingdom M. Harris, Consultant Anesthesiologist J. Simpson, Consultant Anesthesiologist K. Daburn, Research Fellow, Anesthesiology Northern General Hospital, Sheffield, United Kingdom K. Sherry, Consultant Anesthesiologist E. M. Wright, Research Fellow, Anesthesiology Clinic Universitaire de St. Luc, Brussels, Belgium M. Goenen, Professor, Intensive Care J. Jacquet, Attending Physician, Intensive Care Hopital du Bocage, Dijon, France B. Caillard, Professor Anesthesia and Resuscitation J. P. Lancon, Anesthesiologist F. Brennan, Sterling-Winthrop, Guildford, Kingdom M. Main, Sterling-Winthrop, Guildford, Kingdom P. Poole, Statistician, Sterling-Winthrop, rey, United Kingdom
Surrey, United Surrey, Guildford,
United Sur-
Robert 0. Feneck, for The European London,
United
Multicenter
Trial
Kingdom.
Milrinone is a bipyridine methyl carbonitrile analog of amrinone that has been reported to be 10 to 30 times more potent than amrin0ne.l It is a nonglycosidic, noncatecholamine agent with both positive inotropic and vasodilator effects that have been demonstrated in both experimental animals2 and humans.3 The mechanism of action would appear to be by selective and potent inhibition of the cyclic adenosine monophosphate (CAMP)-specific cardiac phosphodiesterase F III. This leads to an increase in cyclic adenosine monophosphate levels, thereby increasing intracellular calcium availability.4 Milrinone has previously been found to be an effective treatment in patients with chronic heart failure, where it has produced a significant fall in pulmonary capillary wedge pressure (PCWP) and increases in cardiac index (C!I).l, 5, 6 There is no data on the use of milrinone in patients recovering from cardiac surgery, but because the proven effects of phosphodiesterase inhibitors have been shown to include enhanced contractility, a reduction in left venReprint requests: Robert Chest Hospital, Banner *See
Appendix
4/O/28163
Group*
0. Feneck, Consultant Anesthesiologist. Road, London E2, United Kingdom.
fbr participating
investigators
and
centers.
London
tricular afterload, and improvement in left ventricular diastolic compliance ,4 there are good theoretic grounds for the assumption that milrinone may be a useful agent in the treatment of low output states in patients recovering from cardiac surgery. We have therefore investigated its use in such patients. PATIENTS
AND
METHODS
Adult patients scheduled to undergo elective cardiac surgery at one of the five participating centers and who gave written informed consent were eligible to enter the study. The trial had the approval of all the local ethical committees concerned and also the approval of the European Ethical Review Committee. All patients were free of significant primary renal or hepatic disease. No patient had a serious hematologic abnormality or an abnormal platelet count. No long-acting cardiovascular medication was given on the morning of surgery, but patients received their routine sedative anesthetic premeditation, and general anesthesia was established in accordance with routine practice in each of the participating centers. Similarly, cardiopulmonary bypass perfusion and surgical techniques were carried out in accordance with local routine practice and protocols. After surgery, all patients were transferred to the intensive care unit. At this time the patients were monitored by continuous display of the ECG and invasive systemic blood pressure. In addition, right atria1 pressure (RAP), mean 1995
1996
Table
Feneck for The European I. Patient
BSA, graft;
+ MVR
Body surface AVR, atria1
Trial
Group
data
No. of patients Age (yr) M/F BSA (m2) Surgery CABG AVR MVR AVR
Multicenter
Low dose
Mid dose
High dose
34 64 26/8 1.86
34 63 20114 1.79
31 62 20/l 1 1.78
27 3 2 2
20 4 9 1
21 5 3 2
area; M/F, male/female; valve replacement; MVR,
CABG, mitral
coronary artery bypass valve replacement.
pulmonary artery pressure (PAP), and PCWP measurements were made by means of a balloon-tipped, flowdirected, pulmonary artery catheter. Cardiac output was measured by the bolus thermodilution technique in accordance with standard practice, and intermittent blood gas analysis was also carried out. All vascular pressures were measured by suitably calibrated, commercially available transducers and displayed on appropriate bedside monitors. In each patient cardiac output measurements were made at a comparable point during the respiratory cycle. Multiple measurements were made at each time and the mean calculated. After surgery all patients had a 12-lead ECG, a chest radiograph, and a platelet count before they received milrinone. Patients were eligible to receive milrinone if they fulfilled the following criteria: (1) stable with regard to ventilation and gas exchange, sedation and analgesia, and fluid and diuretic requirements; (2) low CI, which was defined as a CI of less than 2.51 L/min/m2; (3) adequate cardiac filling pressure, defined as a minimum PCWP of 8 mm Hg but variable for each patient and determined by each investigative team; and (4) hemodynamic stability, which required two consecutive readings of CI and PCWP to be made within a lo-minute period. These measurements had to be such that the lower of the readings was within 15% of the upper reading. In addition, these measurements had to be made within 60 minutes of receiving treatment with milrinone. At these times heart rate (HR), mean systemic arterial pressure (MAP), PAP, and RAP measurements were recorded, in addition to CI and PCWP. Patients were excluded from the study if they demonstrated any evidence of uncontrolled supraventricular tachyarrhythmia, hemodynamically or prognostically significant ventricular extrasystoles, if they had received any long-acting cardiovascular medication, including vasodilators, or if they demonstrated any untoward response to anesthesia or surgery. Milrinone was administered as an initial loading dose of 50 pg/kg in each patient over a lo-minute period. This was followed by a continuous infusion of either 0.375 rglkglmin (low dose), 0.5 wg/kg/min (mid dose), or 0.75 pg/kg/min
American
June 1991 Heart Journal
(high dose). Patients were sequentially allocated to each dosage group in each center. Thus the low-dose group was completed first, then the mid-dose group, and finally the high-dose group. Further measurements of the hemodynamic parameters described were made after 15,30,45, and 60 minutes and after 3, 6, and 12 hours of milrinone therapy. In addition, measurements were made at 2 and 4 hours after milrinone therapy was stopped. Adverse events were recorded in each of the five centers. Data were recorded to establish whether the adverse event was minor or serious, the intensity of the event (i.e., mild, moderate, or severe), the cause of the event (i.e., drug related or not), and what action if any was needed. Statistical analysis of the data was carried out by one-way analysis of variance and paired and unpaired Student t tests as appropriate. RESULTS
Patient data are shown in Table I. As can be seen, there were no significant differences in age, body surface area or weight between the three groups. There were more women in the middle-dose than in the low-dose group, and the middle-dose group also contained a greater number of patients undergoing mitral valve surgery. Baseline hemodynamic data are shown in Table II. There were no significant differences in baseline values for HR, CI, stroke index (SI), PCWP, systemic vascular resistance (SVR), MAP, RAP, PAP, or pulmonary vascular resistance (PVR), and thus the groups were hemodynamically comparable before receiving the drug. Hemodynamic data during and after milrinone treatment are shown in Tables III to VI; the data are expressed as the mean percentage change in each hemodynamic parameter from its baseline value in each of the three groups. Tests of significance are also shown as within-group tests only. As can be seen, after the loading dose there was a rapid and highly significant increase in CI that was well maintained throughout the treatment period in each of the three groups. This increase in CI persisted even after milrinone was discontinued nificantly greater than baseline values
and was sigin each of the
three groups 4 hours after milrinone therapy had been discontinued (Table III). There was a rapid fall in PCWP in each of the three groups
that
was
highly
significant
for
the first
60
minutes of treatment and remained variably significant
in each
of the three
groups
thereafter.
After
treatment, was stopped, PCWP increased in each group. Thus 4 hours after treatment was stopped, PCWP was higher than baseline in the low-dose group, which was similar to baseline in the middle-
Volume
121
Number
6.
Table
Part
Postcardiac
2
II. Baseline hemodynamic data (mean k SEM)
Hg) 11.6 (0.5) 11.9 (0.5) 11.2 (0.5)
15 Min CI (L/m2) LOW Mid High
81 (2.4) 81 (2.3) 83 (3.3)
PCWP Low Mid
SI (ml/m? Low Mid High
23 (1.0) 24 (1.0) 25 (1.5)
High
86 (2.6) 85 (2.5) 81 (2.9)
+40* (4.2) +30* (4.5) +36* (4.9) Hg) -30* (4.7) -34* (4.5) -35* (4.0)
12
hr
Post
+42* (4.9) +34* (4.5) +44* (4.7)
+58* (6.8) +49* (5.1) +66* (6.5)
+44* (6.3) +27* (3.8) +47* (6.6)O
-20*
-15t (7.2) -2o* (4.3) -15t (6.0)
+201 (6.3) -3 (6.7)
(4.7) -25* (4.1) -22* (4.3)
4.X (6.3)
set
$p < 0.01.
Table
IV. Hemodynamic data % Change (mean r SEM)
cm--5) 1876 (102) 1840 (90) 1698 (84) SI (ml) Low
Hg) 8.4 (0.6) 9.0 (0.5) 9.1 (0.5)
Mid
set
cmm5) 202 (12) 224 (19) 184 (16)
dose group, and lower than baseline in the high-dose group. The increase in CI occurred both as a result of an increase in HR and SI (Table IV). These parameters were significantly increased throughout the treatment period and remained variably but significantly increased thereafter. The effects of treatment on MAP were more variable. As can be seen, MAP fell during treatment. However, the effects of milrinone on SVR were more striking; there was a rapid, wellmaintained, and highly significant reduction in SVR in each of the three groups, and this persisted during and after treatment (Table V). Table VI shows the results of hemodynamic measurements from the right side of the heart. As can be seen, there was a fall in RAP immediately after treatment with the loading dose of milrinone in each
15 Min
60 Min
+22.8* (3.5) +14.6*
+25.6* (4.7) +19.1* (4.0) +25.1* (4.3)
+41.8* (5.7) +35.6* (5.5) +42.2* (4.7)
+32.9* (5.2) +20.7* (4.5) +33.5* (6.6)
+13.4* (1.8) +13.6* (2.8) +15.0*
+11.1* (2.9) +ll.Si (3.3) +17.3* (3.6)
+8.7t (3.6) +7.4t (3.2) +11.41: (3.8)
(4.2) High
PVR (dynes Low Mid High
(mm
60 Min
*p < 0.001 tp < 0.05.
Hg)
SVR (dynes Low Mid High RAP (mm Low Mid High
Ill. Hemodynamic data
Table
1.86 (0.06) 1.95 (0.05) 1.98 (0.06)
(beats/min) Low Mid High
MAP (mm Low Mid High
1997
90 Change (mean k SEM)
CI (L/m21 I,OW Mid High PCWP (mm Low Mid High HR
surgery
+19.5* (4.8)
HR (beats/min) Low +14.4* (1.4) Mid +13.9* (1.9) High +14.5* (2.3)
(2.2)
12 hr
Post
*p < 0.001. tp < 0.05. $p < 0.01.
of the three groups. However, this fall was not maintained during the treatment period, and the changes were not significant. Similarly, the effects of treatment on PAP suggest an initial reduction that is variable both within and between the groups and of limited significance. The effects of treatment on PVR were more reproducible, however, in that there was a highly significant reduction in PVR in each of the three groups after 12 hours of milrinone therapy. Table VII shows the data regarding adverse events during this study. As can be seen, there was no difference between the groups regarding the incidence, nature, or severity of the adverse events seen.
1998
Table
Feneck for The European
V.
Hemodynamic
15 Min
60 Min
MAP (mm Low
(2.2) High
-33.3t (2.5)
Trial Group
data 9; Change
Hg) -5.4* (2.5) Mid -11.7t (2.0) High -13.71 (2.2) SVR (dynes set Low -29.5t (2.4) Mid -29.31
Multicenter
Table (mean
-3.9
-8.6’
(2.3)
(3.8)
+2.3 (3.7) -2.6 (3.4) -2.9 (4.9)
-12.6t
(2.3)
cmm5) -27.3t (3.0) -30.1t
(2.5) -35.1t (2.5)
-39.1t (2.9) -39.1t (2.8) -43.2t (2.8)
Hemodynamic
-25.67 (3.8) -22.0t (4.4) -27.3t (4.9)
‘p < 0.05. tp < o.nn1 $p < 0.01.
DISCUSSION
Previous studies have evaluated the effects of milrinone in patients suffering from chronic heart failure.1*5*6 These data provide a valuable insight into the safety and efficacy of milrinone, and in this study we have used the loading dose and maintenance infusion regimens found to be effective in patients with chronic heart failure. Nevertheless, it is clear that there may be substantial differences between patients who have chronic heart failure and those recovering from cardiac surgery. In the latter group, poor left ventricular function may be attributed to poor preoperative left ventricular function and inadequate myocardial protection, thereby leading to a degree of ischemic or reperfusion injury, or a combination of the above factors linked with the depressant effects of anesthesia and cardiopulmonary bypass. Of particular importance may be the effects of hypothermic cardiopulmonary bypass on diastolic function and ventricular compliance that may not have fully recovered during the early postoperative period. In spite of these differences, treatment with milrinone was associated with a rapid and sustained improvement in CI and PCWP in all three patient groups. These data are in agreement with data from patients with chronic heart failure in whom similar effects were also seen. The increase in CI seen in our study was associated with an increase in SI and HR, but this latter increase, although highly significant, was noted more for the statistical reproducibility of the effect than for its magnitude. Other data also suggest that the chro-
15 Min RAP (mm Low
June 1991 Heart Journal
data % Change
Post
-12.6* (3.1)
VI.
t SEM)
12 hr
-9.11 (1.9) -9.4* (3.1)
American
(mean
-+ SEMI
60 Min
12 hr
PUSi
+13.1 (34.4) -9.6 (5.6) -8.4 (6.9)
+46.3 (56.1) -8.7 (4.1) +51.7 (50.5)
+54.4 (42.8) +18.6 (18.1) +72.8 (63.7)
- 10.3t (3.3) -8.7 (3.6) -14.3’ (3.8)
-9.8 (5.7) -14.2t (3.7) -5.4 (4.5)
+8.1
-24.2* (6.9) -5.3 (11.4) -8.1 (7.3)
-36.01 (5.4) -25.7* (9.4 -26.7t (6.9)
Hg)
Mid
-14.8 (11.7) -26.6
High
-22.1
(4.6) (6.5) MAP (mm Low
Hg) -15.6* (2.8) Mid -17.0* (2.6) High -14.7* (3.2) PVR (dynes set Low -25.6* (3.9) Mid -10.1 (6.0) High -7.9 (6.7)
(4.8) +1.1 (5.2) +11.4 (5.6)
cmm5) -28.71
(7.2) -2.3 (13.9) +2.4 (9.3)
*p < 0.01. tp < 0.05. tp < 0.001
Table
VII.
Adverse
Low, 16/34 (47%) 6 Arrhythmias 5 Nausea *Total of seven arrhythmias nonsustained ventricular tivity.
events* Mid, 16/34 (47%) 5 Arrhythmias 4 Nausea deemed tachycardia
drug related; or multifocal
High, 14131 (45%) 6 Arrhythmias 4 Nausea no evidence ventricular
of sustained/ ectopic ac-
notropic effects of milrinone are mild both in absolute terms and in comparison with other treatments.5 We believe that this is an important point, since tachycardia is well known to have the potential for inducing myocardial oxygen imbalance and has also been shown to be the hemodynamic abnormality that correlates most significantly with the development of perioperative ischemic episodes in surgical patients.l Clearly, from our data milrinone is seen to be an effective systemic vasodilator, and indeed this effect could account substantially for the hemodynamic effects seen in this group of patients. However, the effects of the drug on MAP were not marked, and after 12 hours of milrinone therapy, the mean reductions in MAP were between 8 % and 13% in the three groups. These changes were similar in magnitude to those produced by the loading dose only, although
“Ohw Number
121 6, Part
Postcardiac
2
it should be said that during the loading dose period, transient increases in MAP were also noted, which were quite marked in a very small number of patients. As shown above, the effects of milrinone on RAP, PAP, and PVR were not so marked as the other hemodynamic effects seen, although there were highly significant reductions in PVR in all three groups by the end of the treatment period. Clearly the effectiveness of a pulmonary vasodilator drug is substantially affected by the resting tone of the pulmonary vasculature, and while there were pulmonary hypertensive patients in each of the three treatment groups, the baseline data of the three groups did not suggest that pulmonary hypertension is a feature of the groups as a whole. Thus we will need to separate out the pulmonary hypertensive patients and analyze these patients as a further subgroup before we can reach meaningful conclusions about the effects of milrinone on pulmonary hypertensive surgical patients. From the data in Table VII it can be seen that the incidence of adverse effects was similar in the three groups of patients. However, it should be remembered that the incidence of minor adverse effects in patients recovering from cardiac surgery is also common. In particular, the two adverse effects highlighted in Table VII, arrhythmia and nausea, are particularly common in patients recovering from cardiac surgery. It should be emphasized that these were trivial arrhythmias without hemodynamic effects or prognostic significance and that only five of 16 were possibly drug related. Two adverse events were deemed to be serious by the investigators. These were both episodes of atria1 fibrillation, in one patient occurring de novo, and in the other patient occurring as an acceleration of the ventricular rate in prexisting atria1 fibrillation. In both of these cases milrinone treatment was continued for some hours, and the patients’ outcome was not adversely affected. Both episodes of atria1 fibrillation occurred within 5 minutes of administering the loading dose of milrinone in otherwise stable patients and therefore were deemed to be treatment related. In conclusion, our analysis of this data suggests that the effects of milrinone that we have seen in patients with low cardiac output recovering from elective cardiac surgery are very similar to those seen previously in patients who have chronic heart failure. These beneficial effects were well maintained throughout the treatment period, and the incidence of adverse events was acceptably low.
surgery
1999
REFERENCES
JL, Bairn DS. Fein SA. Goldstein RA, LeJemtel TH, Likoff MJ. Efficacy and safety of sustained (48 hr) intravenous
1. Anderson infusions
of milrinone
failure; a multicenter
in patients
with
severe
congestive
heart
study. J Am Co11Cardiol1987;9:711-22.
2. Alousi AA, Stankus GP, Stuart JC, et al. Characterization the cardiotonic effects of milrinone, a new potent cardiac
of
bipyridine, on isolated tissues from several animal species. J Cardiovasc
3.
Pharmacol
1983;5:804.
Ludmer PL, Wright RF, Arnold MO, et al. Separation of the direct myocardial and vasodilator actions of milrinone administration by an intracoronary 1986;73:13&
infusion
technique.
Circulation
Chatterjee K. Newer oral inotropic agents; phosphodiesterase inhibitors. Crit Care Med 1990;18:S34-8. 5. Biddle TL, Benotti JR, Creager MA, et al. Comparison of intravenous milrinone and dibutamine for congestive heart failure secondary to either ischemic or dilated cardiomyopathy. Am J Cardiol 1987;59:1345-50. 4.
6. Benotti
JR, Lawrence
JL, McCue
inetics and pharmacodynamics gestive 7.
heart
failure.
JE, Alpert
of milrinone
Am J Cardiol
JS. Pharmacok-
in chronic con-
1985;56:685-9.
Slogoff S, Keats AS. Does perioperative myocardial ischemia lead to postoperative myocardial infarction? Anesthesiology 1985;62:107-14.
APPENDIX. THE EUROPEAN TRIAL GROUP
MILRINONE
London Chest Hospital, London, United R. 0. Feneck, Consultant Anesthesiologist A. Hill, Research Fellow, Anesthesia R. K. Walesby, Consultant Cardiothoracic
MULTICENTER Kingdom
Surgeon
National Heart Hospital, London, United Kingdom M. Harris, Consultant Anesthesiologist J. Simpson, Consultant Anesthesiologist K. Daburn, Research Fellow, Anesthesiology Northern General Hospital, Sheffield, United Kingdom K. Sherry, Consultant Anesthesiologist E. M. Wright, Research Fellow, Anesthesiology Clinic Universitaire de St. Luc, Brussels, Belgium M. Goenen, Professor, Intensive Care J. Jacquet, Attending Physician, Intensive Care Hopital du Bocage, Dijon, France B. Caillard, Professor Anesthesia and Resuscitation J. P. Lancon, Anesthesiologist F. Brennan, Sterling-Winthrop, Guildford, Kingdom M. Main, Sterling-Winthrop, Guildford, Kingdom P. Poole, Statistician, Sterling-Winthrop, rey, United Kingdom
Surrey, United Surrey, Guildford,
United Sur-
