Secondary Logo

Journal Logo

Cardiovascular Anesthesia

The Effects of Milrinone on Platelets in Patients Undergoing Cardiac Surgery

Kikura, Mutsuhito MD; Lee, Mi K. MD, PhD; Safon, Rebecca A. RN; Bailey, James M. MD, PhD; Levy, Jerrold H. MD

Author Information


Milrinone is a nonglycosidic, nonsympathomimetic drug that increases myocardial cyclic adenosine monophosphate (cAMP) concentration by selective inhibition of cardiac phosphodiesterase fraction III (PDE III) enzymes and increases intracellular calcium delivery, thereby increasing myocardial contractility [1]. This new PDE III inhibitor has beneficial effects on the acute treatment of congestive heart failure [2], and offers an important therapeutic option for ventricular dysfunction in patients undergoing cardiac surgery because of its unique inodilator effects. PDE III enzymes are present not only in cardiac muscle and vascular and bronchial smooth muscle, but also in platelets. In platelets, cAMP generated from adenosine triphosphate by adenyl cyclase serves as an intracellular messenger to inhibit the platelet activation sequence at numerous steps [3]. Since abnormal bleeding after cardiopulmonary bypass (CPB) is most often due to an acute acquired defect in platelets [4], preservation of platelet function is critical to maintaining normal hemostasis in patients undergoing cardiac surgery. Although long-term oral amrinone administration has been reported to produce thrombocytopenia [5-7], no information is available on the acute effects of milrinone on platelet function. Consequently, we designed a controlled prospective study to evaluate the effects of milrinone on platelet number and function in patients undergoing elective cardiac operations.


After Human Investigations Committee approval and informed consent, we studied 27 adult patients electively scheduled for cardiac operations requiring CPB. Exclusion criteria included emergency surgery, history of recurrent ventricular tachycardia or obstructive cardiomyopathy, administration of amrinone within 12 h prior to surgery, abnormal preoperative bleeding time, platelet count less than 100,000/micro Liter, and/or increased serum creatinine (>2.0 mg/dL).

Preoperative medication consisted of intramuscular morphine 0.1 mg/kg, scopolamine 0.2-0.4 mg, and oral diazepam 0.1 mg/kg. Anesthesia was induced with fentanyl 30-75 micro gram/kg or sufentanil 4-8 micro gram/kg, midazolam 0.05-0.1 mg/kg, and vecuronium 0.1-0.2 mg/kg, and the patients were ventilated with inspired 100% oxygen. Patients received fentanyl (total dose 37-105 micro gram/kg) or sufentanil (total dose 5.8-11.5 micro gram/kg), supplemented with midazolam (total dose 0.1-0.4 mg/kg) or diazepam (total dose 0.2-0.5 mg/kg) and/or enflurane (0.5%-2.0%).

After administration of intravenous heparin (400 IU/kg beef lung heparin), aortic and venous cannulas were inserted. Systemic anticoagulation was verified by an activated clotting time (Hemachron 400; International Technidyne, Edison, NJ) >350 s. CPB was conducted using a Cobe membrane lung oxygenator (Cobe, Arvada, CO) and a nonpulsatile flow of 2.2-2.5 L centered dot min-1 centered dot m-2. The circuit was primed with 1500 mL balanced salt solution, 150 mL 15% mannitol, and 500 mL hetastarch. In all patients, hypothermia (23-28 degrees C) and aortic cross-clamping with cold hyperkalemic cardioplegia were used during CPB. For pH management, alpha-stat methodology was used, and the activated clotting time was maintained >350 s. After the primary surgical operation, patients were warmed to a bladder temperature of 36.5-37 degrees C.

Immediately after separation from CPB, patients were randomized to received no milrinone (control group, n = 10), or milrinone (milrinone group, n = 17) at a loading dose of 50 or 75 micro gram/kg immediately prior to separation from CPB followed by 0.5-0.75 micro gram centered dot kg-1 centered dot min-1 continuous infusion for 12-24 h. After separation from CPB, blood remaining in the venous reservoir was reinfused. Protamine sulfate was given after separation from CPB at a ratio of 1 mg:100 U heparin administered to reverse anticoagulation. After obtaining hemostasis, the chest was closed, and patients were transported to the intensive care unit. Any patients who received blood products, including platelets, throughout the study period were excluded.

Perioperative blood samples were obtained from the radial artery catheter for measurements of hematocrit, platelet count (Coulter Counter Trademark; Coulter Electronics Inc., Hialeah, FL), prothrombin time, partial thromboplastin time (ACL Trademark 3000+; Instrumentation Lab., Lexington, MA), platelet aggregation, and thromboelastograph (TEG) before anesthetic induction and 2 and 24 h after separation from CPB. A standardized (i.e., template) bleeding time was measured in the upper extremity using Simplate Trademark II R (Organon Teknika Corp., Durham, NC) before anesthetic induction and 2 and 24 h after separation from CPB. Arterial blood samples for platelet aggregation were collected in 0.1 vol of sodium citrate, and platelet rich plasma (PRP) was prepared by centrifugation at 200g for 10 min at room temperature. Platelet poor plasma was prepared by centrifugation of PRP at 13,000g for 3 min. All equipment used for the blood or platelet plasma was plastic except for the aggregometer tubes, which were untreated glass. Measurement of aggregation was performed by turbidimetric method [8], using a Chrono-log Trademark double-channel aggregometer (model no. 440-VS; Havertown, PA). PRP was incubated for 2 min at 37 degrees C, with stirring, prior to the addition of any stimulating agent. Aggregation studies were performed by exposure of PRP to adenosine diphosphate (ADP) and collagen, and maximum aggregating responses to aggregating agents were defined as the maximum increase in light transmission. The ADP was kept frozen as a stock solution of 1 micro Meter and was diluted at the time of aggregation studies. The concentration of the aggregating agents are expressed as final cuvette concentrations. We prepared variable concentrations of ADP (0.3 to 38.5 micro Meter) and collagen (0.15 to 38.5 micro gram/mL) by dilution of an original stock solution. The concentration of agonist required to produce half-maximum aggregation of platelets, ED50, was calculated for each agonist as follows: After obtaining aggregations at the variable concentrations of ADP and collagen, a computer program (SigmaPlot; Iandell, San Rafael, CA) was used to fit a sigmoidal dose response curve to aggregation responses, and ED50 was determined in each individual's responses. Chest-tube drainage was recorded at the time of arrival in the intensive care unit, and 24 and 48 h after CPB. TEG (Helige Thromboelastograph Trademark; Haemoscope Corporation, Morton Grove, IL) determination was performed for blood coagulability in whole uncitrated blood (360 micro Liter), using disposable plastic cuvettes and pistons. Whole blood was instilled into the cuvette where the pin was lowered, and a thin layer of liquid paraffin was placed on top of the blood sample. The TEG tracing was measured for reaction time (min); initial fibrin formation (min); speed of clot formation, and fibrin cross-linking (alpha); maximum clot strength (maximum amplitude; MA, mm); and clot retraction or lysis (amplitude 60 min after MA; mm). Data are expressed as the mean +/- SD. Statistical analysis was performed using one-way and two-way analysis of variance, followed by the Bonferroni multiple comparison test, and P value less than 0.05 was considered statistically significant.


Demographic data for the control and milrinone groups are described in Table 1. There were no significant differences in age, body weight, height, body surface area, aortic cross-clamp time, time on CPB, and anesthetic time between the groups. In the milrinone group, three patients who received blood products including platelets were excluded from the present study. No abnormal bleeding was observed in any patients after cardiac surgery in this study. Plasma milrinone levels at 2 h after protamine were 152 +/- 42 ng/mL in the milrinone-treated patients.

Table 1
Table 1:
Demographic Data

The changes in platelet counts, bleeding times, and platelet aggregations are shown in Figure 1. There were no significant differences in hematocrit, platelet count, prothrombin time, partial thromboplastin time, bleeding time, ED50 of ADP and collagen-induced platelet aggregation, and TEG measurements between the control and milrinone groups Table 2. In the measurement for TEG, there were no significant changes in the TEG variables from baseline in the control or milrinone groups. In comparison between the control and milrinone groups, there were no significant differences in any coagulation variables measured at baseline and at 2 h and 24 h. There were no significant differences in the 24-h chest tube drainage in the control group (1100 +/- 420 mL) compared to the milrinone-treated group (1055 +/- 570 mL). None of the patients in either group received allogeneic blood transfusions for the first 24 h postoperatively, or required reoperation for bleeding.

Figure 1
Figure 1:
Bleeding times, platelet counts, platelet aggregation to adenosine diphosphate (ADP) and collagen expressed as the effective dose to produce 50% aggregation (ED (50)) at baseline (preoperative) and 2 and 24 h after separation from cardiopulmonary bypass. *P < 0.05 compared to baseline.
Table 2
Table 2:
Changes in Hematologic Variables in the Control and the Milrinone Groups


The present study indicates that intravenous milrinone administration did not cause significant changes in platelet number or function in cardiac surgical patients beyond the adverse effects of cardiac surgery and CPB. Despite the ability of milrinone to increase cAMP in myocardial and vascular smooth muscle [9], there are no adverse effects on platelets. Preservation of platelet function in the first 24 h after CPB is an important consideration, especially in cardiac surgical patients, since quantitative (thrombocytopenia) and qualitative alterations do occur in platelets. Bleeding after CPB is most often due to acquired platelet dysfunction, as the result of interactions between the platelets and nonendothelial surfaces of the extracorporeal circuit, air-blood interfacing, and proteases [4].

In vitro, PDE III inhibitors can increase cAMP concentration in human platelets. Cyclic nucleotides (cAMP) produce platelet dysfunction by three different mechanisms [10]. In platelets, cAMP activates protein kinase to phosphorylate substrate proteins and retards the platelet activation process [3]. Also, cAMP inhibits phospholipase activation, thereby inhibiting the liberation of arachidonic acid from membrane phospholipids [11-13]. Furthermore, cAMP inhibits cyclooxygenase, and thereby inhibits the conversion of arachidonic acid into prostaglandin endoperoxides [14], thus inhibiting the production of thromboxane A2 that is one of the most potent platelet agonists [15]. Third, cAMP stimulates a calcium pump that lowers intracellular calcium levels and impairs calcium-dependent events, such as phospholipase activation and microfilament contraction [16]. We evaluated the effects of milrinone because of the potential effects of phosphodiesterase inhibitors to produce platelet dysfunction.

Amrinone, the first available PDE III inhibitor, has been associated with thrombocytopenia during longterm oral therapy [5-7]. Although the mechanism of amrinone-induced thrombocytopenia is not clear, this is thought to be a nonimmune-mediated toxic effect of amrinone or its metabolities (N-acetylamrinone) on megakaryocytes or platelets, resulting in decreased survival time of the platelet [6,17,18]. Amrinone and milrinone inhibit human platelet aggregation in vitro [19], and milrinone reduces platelet aggregation induced by ADP in animals [20].

In the present study, we noted platelet counts decreased significantly in both groups from the preoperative values at 2 and 24 h after CPB. This reduction in platelet count is likely produced by hemodilution and platelet aggregation induced by platelet-surface interactions during CPB [21,22]. In cardiac surgical patients, significant reductions of platelet counts, even when corrected for dilution based on the hematocrit, persist for 24 h after CPB [23]. There were no significant differences in platelet counts between the control and milrinone groups.

We also observed significant increases in bleeding times from the baseline at 2 and 24 h after CPB in both groups, but there were no differences between control and milrinone-treated patients. Although the bleeding time is a rather crude test, it is one of the only readily available tests to examine platelet function in vivo. Prolongation of bleeding time observed at 2 and 24 h after CPB in this study is due to significant reductions of platelet counts and function during the first 24 h after CPB [24]. We also found a significant reduction in aggregation using standard platelet aggregometry, in all patients, but no difference between milrinone or control patients. We chose ADP and collagen as our agonists, as ADP is a relatively weak agonist, and collagen is a more potent agonist compared to ADP. Using these different agonists, we calculated ED50, to provide a more precise quantification of platelet function than single fixed-dose data previously described [25]. In the present study, both ED50 of ADP and collagen normalized 24 h after CPB, and no significant differences were observed between groups.

We noted that in vitro platelet function measured by aggregation appears to recover more rapidly than does platelet count or bleeding times in the present study. Despite the in vitro platelet aggregation tests, ADP and collagen are only extrinsic agonists to induce aggregation in vitro. Although this test provides the most sensitive modalities for evaluating platelet function in vitro, contribution of platelets for hemostasis in vivo is far more complicated. Therefore, platelet function measured in vitro does not necessarily reflect effects that can be determined in vivo by examining bleeding time. TEG in our study also did not demonstrate any in vitro effects. Recent data has also suggested that the acquired bleeding defect seen after CPB is due in part to the lack of available platelet agonists in vivo [26].

In the present study, we also used the TEG to detect any additional abnormality in the interaction of platelets and the coagulation cascade during milrinone administration. It has been reported that the TEG MA could be a sensitive monitor for the postoperative abnormal bleeding episodes with high sensitivity (85%) [27]. During the period of this study, no significant differences in the TEG MA or any other TEG variables were observed between the control and milrinone groups. This observation was considered to indicate that milrinone administration did not cause any significant platelet-fibrinogen dysfunction.

In summary, intravenous milrinone administration did not cause significant changes in platelet number, platelet function, or viscoelastic properties of blood in cardiac surgical patients beyond the adverse effects of cardiac surgery and CPB.


1. Weishaar RE, Quade MM, Schenden JA, Evans DB. Relationship between inhibition of cardiac muscle phosphodiesterase, changes in cyclic nucleotide levels, and contractile response for Cl-914 and other novel cardiotonics. J Cyclic Nucleotide Protein Phosphor Res 1985;10:551-64.
2. Baim DS, McDowell AV, Cherniles J, et al. Evaluation of a new bipyridine inotropic agent--milrinone--in patients with severe congestive heart failure. N Engl J Med 1983;309:748-56.
3. Campbell FW, Addonizo VP Jr. Platelet function alterations during cardiac surgery. In: Ellison N, Jobes DR, ed. Effective hemostasis in cardiac surgery. Philadelphia: WB Saunders, 1988:93-5.
4. Harker L. Bleeding after cardiopulmonary bypass. N Engl J Med 1986;314:1446-8.
5. Wilmhurst PT, Webb-Peploe MM. Side effects of amrinone therapy. Br Heart J 1983;49:447-51.
6. Ansell J, Tiarks C, McCue J, et al. Amrinone-induced thrombocytopenia. Arch Intern Med 1984;144:949-52.
7. Ross MP, Allen-Webb EM, Pappas JB, McGough EC. Amrinone-associated thrombocytopenia: pharmacokinetic analysis. Clin Pharmacol Ther 1993;53:661-7.
8. Born GVR. Aggregation of blood platelets by adenosine diphosphate and its reversal. Nature 1962;194:927-9.
9. Levy JH, Ramsay J, Bailey JM. Pharmacokinetics and pharmacodynamics of phosphodiesterase-III inhibitors. J Cardiothorac Anesth 1990;4(Suppl 5):7-11.
10. Roevens P, De Chaffoy de Courcelles D. Cyclic AMP-phosphodesterase IIIA1 inhibitors decrease cytosolic Ca2+ concentration and increase the Ca2+ content of intracellular storage sites in human platelets. Biochem Pharmacol 1993;45:2279-82.
11. Gerrard JM, Peller JD, Krick TP, White JG. Cyclic AMP and platelet prostaglandin synthesis. Prostaglandins 1977;14:39-50.
12. Lapetina EG, Schitges A, Chandrabose K. Cyclic adenosine 3 prime,5 prime-monophosphate and prostacyclin inhibit membrane phospholipase activity in platelets. Biochem Biophys Res Commun 1977;76:828-35.
13. Minkes M, Stanford N, Chi MM-Y, et al. Cyclic adenosine 3 prime,5 prime-monophosphate inhibits the availability of arachidonate to prostaglandin synthetase in human platelet suspensions. J Clin Invest 1977;59:449-54.
14. Malmsten C, Gransrom E, Samuelsson B. Cyclic AMP inhibits synthesis of prostaglandin endoperoxide (PGG2) in human platelets. Biochem Biophys Res Commun 1976;68:569-76.
15. Verstraete M, Dejana E, Fuster V, et al. An overview of antiplatelet and antithrombotic drugs. Haemostasis 1985;15:89-99.
16. Kaser-Glanzmann R, Jakabova M, George JN, Luscher EF. Stimulation of calcium uptake in platelet membrane vesicles by adenosine 3 prime,5 prime-cyclic monophosphate and protein kinase. Biochim Biophys Acta 1977;466:429-40.
17. Kinney EL, Ballard JO, Carlin B, Zelis R. Amrinone-mediated thrombocytopenia. Scand J IIaematol 1983;31:376-80.
18. Wilmhurst PT, Al-Hasani SFA, Semple MJ, et al. The effects of amrinone on platelet count, survival and function in patients with congestive cardiac failure. Br J Clin Pharmacol 1984;17:317-24.
19. Pattison A, Astley N, Eason CT, Bonner FW. A comparison of the effects of three positive inotropic agents (amrinone, milrinone and medorinone) on platelet aggregation in human whole blood. Thromb Res 1990;57:909-18.
20. Holbrook M, Coker SJ. Comparison of the effects of isobutylmethylxanthine and milrinone on ischaemia-induced arrhythmias and platelet aggregation in anaesthetized rabbits. Br J Pharmacol 1989;98:318-24.
21. Hope AF, Heyns AD, Lotter MG, et al. Kinetics and sites of sequestration of indium 111-labeled human platelets during cardiopulmonary bypass. J Thorac Cardiovasc Surg 1981;81:880-2.
22. Mammen EF, Koets MH, Washington BC, et al. Hemostasis changes during cardopulmonary bypass surgery. Semin Thromb Hemost 1985;11:281-92.
23. Rinder CS, Johnert J, Rinder HM, et al. Platelet activation and aggregation during cardiopulmonary bypass. Anesthesiology 1991;75:388-93.
24. Holloway DS, Summaria L, Sandesara J, et al. Decreased platelet number and function and increased fibrinolysis contribute to postoperative bleeding in cardiopulmonary bypass patients. Thromb Haemost 1988;59:62-7.
25. Bolt J, Zickmann B, Czeke A, et al. Blood conservation techniques and platelet function in cardiac surgery. Anesthesiology 1991;75:426-32.
26. Kestin AS, Valeri CR, Khuri SF, et al. The platelet function defect of cardiopulmonary bypass. Blood 1993;82:107-17.
27. Spiess BD, Tuman KJ, McCarthy RJ, et al. Thromboelastography as an indicator of post-cardiopulmonary bypass coagulopathies. J Clin Monit 1987;3:25-30.
© 1995 International Anesthesia Research Society