| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
Departments of
*Anesthesiology and
Cardiology, University of Heidelberg, Heidelberg, Germany
Address correspondence and reprint requests to André Gries, MD, Department of Anesthesiology, University of Heidelberg, Im Neuenheimer Feld 110, D-69120 Heidelberg, Germany. Address e-mail to ANDRE_GRIES{at}med.uni-heidelberg.de
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Implications: In a randomized, blinded, placebo-controlled study of patients 24 h after coronary artery bypass grafting, IV administered magnesium inhibited platelet function in vitro and in vivo.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Magnesium also inhibits platelet function in vitro (5) and in vivo (6–8). IV administered magnesium reportedly prolongs the bleeding time in pregnant women and inhibits platelet aggregation in healthy volunteers (6,7), perhaps by inhibiting fibrinogen binding to the platelet glycoprotein (GP) IIb/IIIa receptor on the platelet surface (8).
It is well known that extracorporeal circulation leads to thrombocytopenia and platelet dysfunction in patients undergoing cardiac surgery (9). Impaired platelet function can result in an increase in postoperative bleeding episodes, which may require transfusion of blood or plasma products. Thus, further inhibition of platelet function by drugs with potential antiplatelet properties, such as magnesium, may be hazardous in these patients.
Whether magnesium also inhibits platelet function after cardiac surgery has not been investigated. We therefore studied the effect of magnesium on platelet function in patients after coronary revascularization in a randomized, blinded, and placebo-controlled trial.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
After incubation, ADP- (final concentration 5 µM) and collagen- (final concentration 0.19 mg/mL) induced platelet aggregation was measured in a four-channel platelet amplifier (PAP-4TM; Biodata Corporation, Hartboro, PA) at 37°C according to the method described by Born (10). Aggregation was quantified by measuring the maximal extent of light transmission (maximal aggregation, measured as a percentage).
To determine the effect of magnesium on platelet P-selectin expression, PRP was diluted in HEPES solution and in Tyrode's buffer (Roth, Karlsruhe, Germany) to measure fibrinogen binding to the GP IIb/IIIa receptor. There were no centrifugation or vortexing steps; therefore, in vitro platelet activation was avoided. P-selectin expression and fibrinogen binding were measured after stimulation with ADP (final concentration 2 µM). The samples were incubated with specific antibodies—namely, CD62P CloneAC1.2 (Becton Dickinson, Heidelberg, Germany)—to determine P-selectin expression, and anti-human fibrinogen fluorescein isothiocyanate-conjugated chicken antibody (Biopool, Umeå, Sweden) to determine fibrinogen binding for 20 min in the dark at 22°C. After fixation with CellfixTM (Becton Dickinson), the samples were measured in a flow cytometer within 6 h. The platelet population was analyzed at a low flow rate and was identified on the basis of forward and sideways scatter characteristics. For each sample, 10,000 platelets were collected. Data were analyzed using LYSIS II software (FACScanTM; Becton Dickinson) (11).
To rule out a potential platelet inhibitory role of ascorbate as administered with the study drug, three different drugs containing magnesium were investigated in three additional subjects. Equivalent amounts (in mmol) of commercially supplied magnesium ascorbate (MagnorbinTM), magnesium aspartate (MagnesiocardTM; Verla, Tutzing, Germany), and magnesium sulfate (Magnesium VerlaTM; Verla) 1, 5, and 10 mM were added to PRP as described above.
Magnesium In Vivo
After approval by the ethics committee and informed consent from the patients had been obtained, 46 male patients (64 ± 4 yr old) undergoing coronary revascularization were additionally investigated. No patients had taken any drugs known to affect platelet function during the 8 days before surgery. Anesthesia technique was comparable in all patients; we used only anesthetics known not to affect platelet function (12). After sternotomy, patients were anticoagulated with heparin to an activated clotting time >400 s and cooled to 30°C during CPB using a membrane oxygenator (Maxima Plus PRFTM; Medtronic, Kerkrade, The Netherlands). Protamine was later administered to antagonize the effect of heparin, whereas aprotinin or other antifibrinolytics were not administered. After surgery, all patients were transferred to the intensive care unit.
On the first postoperative day, all patients were reevaluated to determine that surgery and postoperative care had been uneventful. Patients who had received catecholamines other than dobutamine at a maximal rate of 3 µg · kg-1 · min-1 (n = 2), nitrates (n = 2), any other platelet-affecting drugs (heparin, n = 1); and patients who were classified as not stable enough to be transferred to the normal ward in the next 6 h (increased chest tube drainage, requirement of >2 U of packed red cells in the postoperative period, n = 2) were excluded from the study. The remaining 39 patients were randomly allocated to either the magnesium (n = 19) or the control group (n = 20). Twenty-four hours after the patients' arrival in the ICU (PO24), 2 g of MagnorbinTM (equivalent to 5.4 mmol magnesium and 1331 mg of magnesium sulfate) was administered IV over a period of 15 min in the magnesium group, whereas saline was given in the control group. As described above, arterial blood samples for laboratory investigations were withdrawn before anesthesia induction (PRE), 24 h after surgery just before initiating the magnesium or saline infusion (PO24), and 30 min after magnesium or saline administration (postinfusion; PINF).
Using a platelet function analyzer (PFA 100TM; Dade, Unterschleißheim, Germany), the in vitro bleeding time was determined according to the method described by Kratzer and Born (13). Briefly, at a constant negative pressure, samples of citrated whole blood (800 µL) are suctioned through a small capillary and a filter membrane with a diameter of 150 µm. The filter membrane is covered with collagen and soaked with epinephrine. During movement through the capillary, platelets adhere and aggregate at the filter membrane, thus diminishing blood flow until it stops. The total time of the blood flow is designated as the in vitro bleeding time and is measured electronically (13).
As described above, whole blood (8 mL) was centrifuged at 500g for 5 min and subsequently at 2500g for 10 min to prepare PRP and PPP. Adding PPP, PRP was adjusted to a platelet count of 300/nL, and ADP and collagen-induced platelet aggregation was measured.
To determine P-selectin expression and fibrinogen binding to the platelet membrane, whole blood samples (20 µL) were measured as described above after activation with ADP (final concentration 2 µM).
At all sampling time points, the serum magnesium concentration was measured using a specific spectrophotometric assay. In addition, serum calcium concentration, platelet count, hematocrit, fibrinogen levels, antithrombin III, and partial thromboplastin time were determined at time points PRE and PO24. Furthermore, requirement of transfusion of blood products during the study period was documented.
In patients already prone to coagulopathy, platelets after CPB may be more sensitive to the antiplatelet effect of magnesium; therefore, five additional patients (57 ± 12 yr old) undergoing major intraabdominal surgery were studied. After administration of 5.4 mmol magnesium, the in vitro bleeding time, platelet aggregation, P-selectin expression, and fibrinogen binding were investigated according to our study protocol as described above.
To determine platelet function, each test was performed in duplicate, and the mean value of both measurements was recorded. All results are given as mean ± SD. Statistical analysis was performed using one-way analysis of variance for repeated measurements, followed by Scheffé's test to demonstrate changes in bleeding time, platelet aggregation, P-selectin expression, and fibrinogen binding, and to analyze the differences between the two groups. A P value of <0.05 was considered to be statistically significant.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
|
Magnesium In Vivo
Substitution of blood products was comparable in both groups and showed no significant differences. During the study period (PRE-PINF), 927 ± 484 mL (15 patients) of packed red cells and 1022 ± 1029 mL (9 patients) of fresh-frozen plasma had been administered in the control group, and 974 ± 627 mL (13 patients) of packed red cells and 1055 ± 607 mL (10 patients) of fresh-frozen plasma had been administered in the magnesium group. Furthermore, in both groups, dobutamine was administered at a similar infusion rate (3.0 ± 1.9 µg · kg-1 · min-1 in the magnesium group versus 3.0 ± 1.3 µg · kg-1 · min-1 in the control group) and in a similar number of patients (14 vs 11).
Serum magnesium levels were comparable in both groups at PRE: 0.86 ± 0.15 mmol/L in the control group and 0.82 ± 0.17 mmol/L in the magnesium group (not significantly different; NS). After coronary revascularization in both the control (0.74 ± 0.15 mmol/L) and magnesium groups (0.71 ± 0.09 mmol/L), serum magnesium was significantly decreased by 14% at PO24 (P < 0.05 versus PRE). The administration of magnesium significantly increased (18%) magnesium levels to 0.84 ± 0.11 mmol/L in the magnesium group versus 0.75 ± 0.04 mmol/L in the control group (P < 0.01) at PINF and thereby restored magnesium levels to levels that did not differ from the PRE values.
Compared with PRE, the in vitro bleeding time was prolonged, and ADP- and collagen-induced platelet aggregation were inhibited, in both groups after CABG and CPB at PO24. In contrast to the control group, magnesium administration further increased bleeding time and inhibited platelet aggregation at PINF. Similarly, the magnesium infusion also affected both ADP-activated P-selectin expression and fibrinogen binding to the GP IIb/IIIa receptor, which were inhibited compared with the control group at PINF (Table 2).
|
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
After cardiac surgery and CPB, inhibition of ADP- and collagen-induced platelet aggregation and prolongation of the in vitro bleeding time without complete recovery in the postoperative period reflect platelet dysfunction (14–16). Moreover, increases in P-selectin expression before CPB, finally resulting in a loss of functional platelets afterward, and inhibition of GP IIb/IIIa receptor expression have been described by others (17,18). The loss of functional platelets is reflected by a decrease in platelet count after CPB (9,18). Thus, our PO24 data characterize the typical postoperative platelet function state in patients after CABG and CPB.
After CPB, hypomagnesemia is common: 71% of patients are hypomagnesemic after CABG and aortic valve replacement (1,19). Moreover, postoperative hypomagnesemia is associated with prolonged ventilatory support and an increased incidence of cardiac arrhythmias (1,2). The administration of magnesium after CPB reduces the incidence of supraventricular and ventricular arrhythmias and increases the cardiac index and left ventricular stroke work index (2,3). For these reasons, magnesium therapy is regularly considered in patients after CABG and CPB.
Besides these cardiac effects, magnesium inhibits platelet function. In one study, dose-dependent inhibition of platelet aggregation, platelet thromboxane A2 synthesis, and ß-thromboglobulin release from the alpha granula were observed at 0.5–8.0 mM magnesium in vitro, and these effects were significant at clinically relevant concentrations (20). Complete inhibition of platelet aggregation at 10 mM magnesium was reported by others (21). In addition, inhibition of platelet aggregation, reduction of platelet adenosine triphosphate release, and inhibition of thromboxane A2 synthesis were demonstrated in human platelets, PRP, whole blood, and washed platelets (5). Inhibition of fibrinogen-mediated aggregation, 50% inhibition of platelet adhesion at 2 mM magnesium, and inhibition of platelet fibrinogen binding and P-selectin expression with a 50% inhibition at 3 mM magnesium have been observed in washed platelets (8). There are no data, however, on the in vitro effect of magnesium on platelet function in patients undergoing cardiac surgery. We observed a dose-dependent inhibition of ADP- and collagen-induced platelet aggregation and inhibition of fibrinogen binding to the GP IIb/IIIa receptor. Moreover, because platelet function was already inhibited at PO24 compared with PRE, our findings suggest that impaired post-CPB platelet function can be further influenced by magnesium.
This suggestion was strengthened by our in vivo findings. The administration of 5.4 mmol magnesium, which restored the serum magnesium level to presurgical levels, led to a 20% prolongation of the in vitro bleeding time, 11% inhibition of ADP-induced platelet aggregation, and 17% inhibition of collagen-induced platelet aggregation in patients 24 h after CABG. In accordance with these findings, a 48% prolongation of the bleeding time and a 40% inhibition of ADP-induced platelet aggregation were observed in healthy volunteers after the administration of 8 mmol MgSO4 (7,8). Furthermore, magnesium therapy also led to a significant prolongation of the bleeding time in pregnant women (6,22). Animal studies further support these findings (23).
Compared with the findings in the literature, the prolongation of bleeding time and inhibition of aggregation were reduced in the present study. This difference may be explained by the smaller amount of magnesium administered. In the present study, 5.4 mmol magnesium led to an increase in the serum magnesium concentration from 0.71 mmol/L to 0.84 mmol/L in the cardiac surgery group and from 0.81 mmol/L to 1.05 mmol/L in the noncardiac surgery group. In comparison, 8 mmol magnesium increased the serum magnesium concentration to 1.2–1.5 mmol/L in normomagnesemic volunteers, who most likely had a normal platelet count at this time (7,8). In our cardiac surgical patients, platelet function was already inhibited before magnesium administration, probably because they had undergone CPB (9,16). Considering the results from the noncardiac surgery group, magnesium had no effect on platelet function (24), we suggest that platelets in hypomagnesemic post-CPB patients may be more sensitive to rapidly infused magnesium than platelets in normal volunteers.
In the present study, we also observed inhibition of ADP-activated P-selectin expression after magnesium administration. This reflects inhibition of platelet function because, in activated platelets, P-selectin is translocated to the platelet surface from alpha granules. Magnesium also led to inhibition in fibrinogen binding to the GP IIb/IIIa receptor. Although data on the effect of magnesium on P-selectin expression and fibrinogen binding are limited, inhibition of ADP-activated P-selectin expression and fibrinogen binding have been reported in healthy volunteers, which is in accordance with our findings (8).
Moreover, inhibition of fibrinogen binding to the GP IIb/IIIa receptor, which is required for platelet adhesion and aggregation (25), may explain, at least in part, the prolongation of the bleeding time and inhibition of platelet aggregation (21). In contrast, other authors postulate a decrease in intracellular cAMP levels and Ca2+ influx (5,26) and inhibition of cyclooxygenase and lipoxygenase pathways (5), finally leading to inhibition of platelet function. Hypomagnesemia has been demonstrated to impair the release of nitric oxide from coronary endothelium (27), whereas magnesium administration leads to a transient increase in the release of nitric oxide (7). This suggests that platelet inhibition may also be mediated by nitric oxide, which inhibits platelet function in vitro and in vivo (11). Nevertheless, further studies using flow-cytometric methods are required to investigate the effect of magnesium on platelet function.
Regarding the study protocol, magnesium was administered late in the postoperative course and to the healthiest patients. Nevertheless, considering the increased blood loss observed at delivery after magnesium therapy in pregnant women (22), our in vitro findings suggesting that platelets post-CPB are more sensitive to the antiplatelet effect of magnesium, and that magnesium therapy in patients after CPB has been claimed to be harmful by others (28–31), in the present in vivo study, only patients with uneventful surgery and postoperative care were further investigated at PO24. We are not able to state with certainty that the risk of magnesium administration would outweigh its potential beneficial effects. However, this selection was performed, and the time of magnesium administration was chosen, as a safety precaution. We believe that this was appropriate because there are no additional data.
Antioxidants (vitamin E, betacarotene, vitamin C, and others) have been suggested to influence leukocyte adhesion to vascular endothelium and platelet-leukocyte interaction (32), and daily supplementation has been suggested to be beneficial in selected patient populations (32–34). One may assume that platelets were also influenced by ascorbate in the present study because magnesium ascorbate was used in vitro and in vivo. However, data for a significant platelet inhibitory effect of ascorbic acid alone remain controversial (35,36). In addition, when three different agents containing magnesium were investigated (magnesium ascorbate, aspartate, and sulfate) comparable platelet inhibitory properties were observed in vitro. Therefore, it seems unlikely that ascorbate, as administered with the study drug, significantly contributed to platelet inhibition in the present study.
In both groups, there were no bleeding complications, and the requirements of packed red blood cells and fresh-frozen plasma were similar. The incidence of myocardial infarction and arrhythmias was not investigated. In addition, laboratory measurements of platelet dysfunction may be oversensitive. Although recent studies provide evidence for the predictive value of platelet function tests (37,38), a definite threshold for the increase in clinically apparent bleeding episodes or blood loss cannot be provided. Because any clinical effect of magnesium was absent in the present study, care must be taken in extrapolating our findings to the clinical setting. However, because other investigators may administer larger doses to patients with co-existing bleeding disorders (28,39,40), our laboratory findings may reach clinical significance and provide a rational basis for investigating the antiplatelet effect of magnesium in further clinical studies. In a recent study, 14 mmol/70 kg magnesium sulfate was administered before anesthesia induction in patients undergoing CPB and CABG. The perioperative blood loss and transfusion requirement would have been interesting, but were unfortunately not reported (41).
In conclusion, the results from the present study provide new evidence of an inhibitory effect of magnesium on platelet function. This antiplatelet effect may be a factor in the attenuation of the mortality rate in patients treated with magnesium after acute myocardial infarction (4). Conversely, in patients with preexisting bleeding disorders, this additional platelet inhibition may not be beneficial.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
