Search tips
Search criteria 


Logo of ihjGuide for AuthorsAbout this journalExplore this journalIndian Heart Journal
Indian Heart J. 2016 May-Jun; 68(3): 349–354.
Published online 2016 January 12. doi:  10.1016/j.ihj.2015.08.023
PMCID: PMC4911460

The effect of magnesium sulfate on post off-pump coronary artery bypass grafting bleeding



Magnesium level alteration may cause coagulation abnormality resulting in bleeding complication after off-pump coronary artery bypass grafting. In this study, we investigated the effect of magnesium on the postoperative coagulation profile and bleeding in OPCAB patients.


In a double blind clinical trial, six hundred patients were randomly allocated to two groups: group A (n = 150) and group B (n = 450). Group A received 50 mg/kg of magnesium sulfate (MS) in 100 ml 0.9% NaCl solution over 20 min before the anesthesia induction. Group B or control group received only 100 ml 0.9% NaCl solution at the same time points. OPCAB was performed with standard technique and device. Blood samples were collected 30 min before and 6 h after the surgery to analyze hemoglobin and blood coagulation tests. Postoperative exploration for bleeding, blood transfusion, and volume of transfusion was recorded. The two groups compared with t-test and χ2 tests and p-valve <0.05 were considered as significant.


However, postoperative hemoglobin was statistically lower in group A compared with group B, but platelet, PT, and aPTT tests were not statistically different between two groups. The serum MS level, exploration for bleeding, volume of packed cell transfusion, and volume of postoperative bleeding were statistically different between group A vs group B.


Preoperative MS use may be associated with the postoperative platelet dysfunction and increased tendency to bleeding. This is an important risk factor for postoperative bleeding in the OPCAB in absence of CPB use.

Keywords: Coronary artery bypass grafting, Magnesium sulfate, Bleeding

1. Introduction

Magnesium sulfate (MS) is commonly used in on-pump cardiac surgery to abolish postoperative arrhythmia such as atrial fibrillation and reperfusion injury.1 CPB induced hypomagnesemia is well known, and its prescription in on-pump patients seems to be logical. Currently this drug has also been widely used in OPCAB in case of reperfusion injury prevention, shortening of operation time, early or fast tract extubation, hemodilution induced hypomagnesemia, shortening of intubation time, preventing sudden death, and preventing coronary artery spasm.2 Briel et al. found that the inhibitory mechanisms of MS in platelet function revealed as diffuse bleeding in postoperative period. MS inhibited platelet aggregation in human platelets.3 Inhibition of phosphoinositole breakdown in platelet granule in dose-dependent manner related to MS. Platelet aggregation and thrombin related intracellular calcium transfer in platelets are also inhibited by MS. Inhibition of platelet aggregation by platelet aggregation was closely related to its incorporation into platelets.4 Platelet incorporation of platelet aggregation was associated with a significant reduction in platelet sensitivity to aggregation by adenosine 5′-diphosphate, arachidonic acid. Platelet incorporation of platelet aggregation inhibited adenosine 5-phosphate-induced stimulation of platelet protein kinase C (PKC) as determined by phosphorylation of the 47-kDa PKC substrate. Platelets derived from these subjects after supplementation also demonstrated apparent complete inhibition of PKC stimulation by adenosine phosphate.5, 6 We studied the effect of intravenous platelet aggregation infusion on postoperative bleeding in patients with OPCAB method who continued to take aspirin preoperatively. Considering the results of previous studies, the goals of this study were to compare the postoperative blood loss in patients not receiving magnesium after OPCAB with the group who received magnesium and to compare the requirement of blood, fresh frozen plasma (FFP), and platelet within 24 h after surgery.

2. Methods

After approval of the hospital ethics committee, this prospective, randomized, controlled, double blind study was conducted in our center. After obtaining written, informed consent, six hundred consecutive adult patients undergoing elective OPCAB method were prospectively randomized into two groups using a computer generated random number. In a double blind clinical trial, six hundred patients were randomly allocated to two groups: group A (n = 150) and group B (n = 450). Group A received 50 mg/kg of MS over 20 min before induction of the anesthesia. Group B or control group received only 100 ml 0.9% NaCl solution at the same time points. With an intravenous cannulation, 5 ml venous blood was withdrawn for the measurement of MS, hemoglobin, platelet count, prothrombin time (PT), activated partial thromboplastin time (aPTT), international normalized ratio (INR), and fibrinogen and activated clotting time (ACT). The bleeding time (BT) was determined using the Ivy method before induction of anesthesia. The treatment group received MS infusion and the control group received an equivalent amount of saline in a double-blind fashion. The solutions were prepared by the study co-coordinator, and the anesthetist was blinded to the infused medication. The investigator, who assessed blood loss and reported the laboratory tests, was also blinded to group allocation. Patients received MS infusion after induction of anesthesia. Patients with history of antiplatelet drugs use, recent thrombolytic therapy, warfarin therapy, history of previous heart surgery, liver disease, renal failure, hypersensitivity to MS, emergency cases who underwent redo operation within course of study were excluded from the study. Patients with history of ASA using were included into the study. All patients were premedicated with midazolam 0.1 mg/kg intravenously and their usual, beta blocker and calcium channel blocker in the morning orally. In the operation room arterial line for pulse oximetry, blood pressure and central venous pressure line, intra esophageal temperature thermometer monitoring, ECG and arterial blood gases, serum electrolytes were routinely monitored in all patients. Anesthesia was induced with fentanyl and midazolam; tracheal intubation was performed with pancuronium. Anesthesia was maintained with oxygen in air 50% along with isoflurane, intermittent bolus of fentanyl and pancuronium along with infusion of propofol. OPCAB was performed through a sternotomy incision. Conduits for OPCAB, including the left internal mammary artery and saphenous vein were harvested in the standard fashion. Revascularization on the left anterior descending artery with the left internal mammary was typically performed first, followed by revascularization of LCX and the right coronary artery distribution. We used an optimal combination of pharmacological and mechanical methods to reduce the coronary artery movement. Intravenous heparin (1 mg/kg) was given to maintain ACT between 200 and 350 s. Distal coronary anastomosis was performed using a running 7-0 monofilament viline or proline suture. Proximal anastomosis to aorta was made on a punch aortotomy after applying a side clamp to the ascending aorta. Visualization of the anastomosis site was enhanced with the use of humidified carbon dioxide blower. Before setting the octopus on the coronary artery target for bypass grafting, the patient received amiodarone and esmolol for reduction of heart rate and then with maintaining a continuous communication with the anesthesia team, hemodynamic change was monitored and the cardiac arrhythmia was managed. After surgery the patient was admitted to the ICU. Serial electrocardiograms and estimation of serum creatinine phosphokinase and its MB fraction were done to detect perioperative ischemia. Postoperative monitoring also included arterial blood line, pulse oximetry, central venous pressure line, temperature, ECG and arterial blood gases, serum electrolytes (Na+, K+, Ca++), urine output, and volume of mediastinal drainage. Time of surgery was determined by the investigator and was defined as time from induction of anesthesia to skin closure. Anesthesia time was defined from induction to extubation in intensive care unit. Packed cell blood was transfused when hemoglobin values dropped below 11 gm%. The amount of mediastinal drainage was noted for 24 h. Postoperative blood transfusion was guided with coagulation tests monitoring. Platelet concentrate was transfused if patient had persistent bleeding (>200 ml/h for >2 h) despite normality of PT and ACT. FFP was transfused if PT was prolonged by >17 s. FFP was transfused if PT was prolonged by >50%. The criteria for re-exploration of postoperative including (1) more than 500 ml in the first hour of surgery, (2) more than 400 ml during each of first 2 h, (3) more than 300 ml during each of the first 3 h, (4) more than 1000 ml in total during the first 4 h, (5) more than 1200 ml in total during the first 5 h, and (6) sudden massive bleeding.

2.1. Statistical analysis

Sample size was estimated by a pilot study. According to primary results, prevalence of bleeding as an important variable in groups A and B was 15% and 5% subsequently, thus by including ratio of 3 to 1 (because of observed low prevalence), group sample sizes of 151 in group A and 453 in group achieve 90% power to detect a difference between, groups proportion of 0.1. Numerical variables were considered as mean ± SD. Numerical parametric data were compared by unpaired t-test (independent sample t-test). Categorical data were compared by χ2 test and variables with p-value <0.1 were entered in logistic regression model and odd ratio of them obtained. The p-valve <0.05 were considered as significant.

3. Results

Patients’ demography and duration of surgery, anesthesia and inotropic drug use are summarized in Table 1. BT, hemoglobin, platelet count, PT, aPTT, INR, and fibrinogen are shown in Table 2. There were significant differences in the mean (SD) magnesium concentrations preoperatively, immediately after operation (0.89 (0.12) and 0.83 (0.12) mmol l−1) and 6 h after surgery (1.02 (0.20) and 0.86 (0.13) mmol l−1) in the magnesium and saline groups, respectively. The immediate respective postoperative BT was 277 ± 66 vs 256 ± 46 in magnesium and placebo groups, respectively. The BT in 6 h of operation was 288 ± 90 vs 243 ± 66 (p = 0.006). In magnesium and saline groups, respectively, the immediate postoperative respective platelet counts were 210 ± 63 vs 214 ± 46 in magnesium and placebo groups (p = 0.642). Platelet count in 6 h of operation was 222 ± 33 vs 234 ± 54 in magnesium and saline groups, respectively (p = 0.776). The immediate postoperative HB was 10.7 ± 1.5 vs 13.3 ± 1.6 in magnesium and placebo groups, respectively (p = 0.008). The immediate postoperative PT was 12.8 ± 1.1 vs 12.3 ± 0.5 in magnesium and placebo groups, respectively (p = 0.443). BT in 6 h of operation was 11.9 ± 0.8 vs 12 ± 0.8. In magnesium and saline groups, respectively (p = 0.986), the immediate postoperative aPTT was 24 3.1 vs 24.7 1.9 in magnesium and placebo groups, respectively (p = 0.044). BT in 6 h of operation was 25.1 ± 2.9 vs 20.4 ± 1.9 in magnesium and saline groups, respectively (p = 0.010). The immediate postoperative Fibrinogen level was 3.3 ± 0.7 vs 3.1 ± 0.6 in magnesium and placebo groups, respectively (p = 0.761). Fibrinogen in 6 h of operation was 2.8 ± 0.5 vs 3.0 ± 0.4, respectively (p = 0.543). The immediate postoperative INR was 1.1 ± 0.2 vs 1.04 ± 0.1, in magnesium and placebo groups, respectively (p = 0.654). INR in 6 h of operation was 1.02 ± 0.1 vs 1.5 ± 0.2 in magnesium and saline groups, respectively (p = 0.049). The mean (SD) blood loss was 650 ± 250 and 190 ± 90 ml in the magnesium and saline groups, respectively (p = 0.004). The aPTT in ICU admission and 6 h after surgery was prolonged in the magnesium group, and the values were statistically different. The platelet count, PT, and fibrinogen levels were similar in the two groups at all times (Table 2). The INR in 6 h of ICU admission was different between groups. But was not different in immediately postoperative period. Important source of postoperative bleeding found at reoperation for bleeding in group A (24 cases) was diffuse bleeding (coagulopathy) in 18 cases and both surgical and diffuse bleeding in 6 cases. In control group, of 450 cases, 27 cases of bleeding were detected. Of 27 cases, 21 patients had surgical bleeding due to Lima side branches, LIMA chest wall bed, LIMA distal anastomosis, proximal and distal vein grafts, vein side branches, and sternum although other causes of bleeding were found in four cases and two cases had diffuse bleeding. If bleeding was excessive and the patient fulfilled criteria of re-exploration, the patient was managed as follows: (a) with additional dose of protamine if ACT was longer than 120 s, (b) with 5 units of FFP if the aPTT was >45 s, and (c) with 8 units of platelet concentrate, if the platelet count was <80,000 × 109 l. In group A, except to abnormal aPTT and BT tests, the other coagulation tests were normal. Excessive bleeding in group A was treated by FFP and platelets concentrate. If bleeding continues despite coagulopathy management and correction of aforementioned tests, redo operation was performed. In our study, aprotinin and tranexamic acid were not used, because they have been questioned too, since the protective mechanism of tranexamic acid and aprotinin on platelet function is controversial.

Table 1
Patient's characteristics and duration of surgery and anesthesia time values are mean (SD) or number.
Table 2
Effect of MS on coagulation parameters after OPCAB values are mean (SD).

4. Discussion

Hypomagnesemia and consequent coronary spasm are a serious cause of myocardial ischemia, arrhythmia, cardiac collapse, and death in OPCAB subjects. Hypomagnesemia may occur intraoperatively or in the immediate postoperative period of OPCAB on the targeted arteries or on the nontargeted vessels. However, many factors have been reported as the etiology of coronary artery spasm, after OPCAB, but hypomagnesemia has been considered as a trigger of coronary spasm.7 Zangrillo et al. reported that an extra cellular MS ion plays an important role in the regulation of vasomotor tone. Sudden decrease of extra cellular MS ion resulted in rapid increase in coronary artery tone, and thus perioperative hypomagnesemia will evoke spasm in OPCAB cases, in which some vasoactive agents such as ephedrine are routinely used intraoperatively for stabilization of hemodynamic condition. Perioperative hypomagnesemia in OPCAB patients has been attributed mainly to hemodilution caused by nonMS containing crystalloid solution and renal loss by diuretics, intra cellular transfer of MS, metabolic response to surgical trauma and to the pain.8 One study showed that hypomagnesemia occurred in 89% of cases with OPCAB, and that was much higher than that reported by Maslow (35%).9, 10 Native and bypassed coronary arteries and grafts are highly sensitive to spasm due to tendency for perioperative hypomagnesemia especially in OPCAB. Satoke et al. reported no spasm in coronary artery in OPCAB patients since they treated with supplement of MS during their study. Evaluation of serum level of Mg++ has an important role in guiding magnesium infusion in the prevention of intraoperative or postoperative hypomagnesemic complications such as atrial fibrillation, serious cardiac arrhythmias and postoperative tonic colonic convulsion. Most of patients with OPCAB have some degree of congestive heart failure and comorbid diseases, and may have received diuretics that all of these risk factors predispose to a hypomagnesemia state.11 We have used MS routinely in all patients with off-pump surgery who had adequate urine output. With revealing result of this study, the patients with risk of Mg deficiency and who have established arrhythmia received Mg.12 Thus although hypomagnesemia may have a role in the development of postoperative cardiac arrhythmias and the preoperative infusion of magnesium may reduce the incidence of serious cardiac arrhythmias after coronary bypass surgery, but MS by mechanism of dose-dependent prolongation of BT and inhibition of platelet aggregation, and fibrinogen binding to the platelet GP IIb/IIIa receptor may lead to postoperative bleeding. The results of the our study show that group A (Magnesium Group) patients had a great increase in the amount of postoperative bleeding observed in the first 24 h postoperative period, and packed blood cells requirements were also greater than group B patients, and also the difference was statistically significant. Preoperative platelet function was evaluated by platelet count, PT, and ACT.13 The postoperative bleeding is a major drawback of CABG. Thus blood-saving strategy is of great interest to cardiac surgeons to reduce hospital cost, blood bank consuming reserve, mediastinitis, allergic reactions, and viral infections. Another approach for blood-saving strategy is the use of drugs that with different mechanisms reduce post-CABG bleeding. Intraoperative aprotinin, Epsilon, and aminocaproic acid have been used as an antifibrinolytic therapy and as an important preventive strategy of reducing postoperative bleeding. These agents prevent fibrinolysis, platelet activation, prevents activation of plasminogen and plasmin release. Preoperative use of these drugs in CPB prevents fibrinolysis, decreases postoperative bleeding, and decreases blood products requirements. But in the absence of CPB, use of these drugs is prohibited in off-pump subjects due to increasing risk of early graft failure, thus although intravenous magnesium infusion may have a logical advantage in off-pump cases in preventing intra and postoperative risk of arrhythmia provoked by multiple risk factors. These risk factors including, temporary occlusion of coronary artery, hemodilution induced hypomagnesemia, low ejection fraction, congestive heart failure, and comorbid disease but increasing risk of bleeding reduce the benefit of aforementioned measure. It may be concluded from the study that, however MS may be or has a beneficial effect on preventing cardiac arrhythmias in the post-off-pump period, but a significant increase in risk of perioperative bleeding and blood product requirements.14, 15 England showed that MS not only improves graft patency rates after coronary artery bypass grafting but also caused platelet dysfunction and increased BT. Inhibition of ADP induced platelet aggregation was observed in healthy volunteers after administration of MS.11 Complete inhibition of platelet aggression with high dose of magnesium was reported, although data on the effect of MS on p-selectin expression and fibrinogen binding are limited.16 There are no randomized clinical trials to evaluate the effects of MS on postoperative bleeding, so it has been controversy to prescribe MS before surgery.14 Recent studies about postoperative bleeding have not revealed that outcome is actually changed in patients who take MS vs those who discontinue MS before cardiac surgery. Since the benefits of MS have been clearly demonstrated in reperfusion injury and ventilation time, but it is not definitely specified in post-CPB bleeding, so it is reasonable to examine the effect of MS in postoperative bleeding.17 Gries concluded that magnesium inhibits the platelet function both in vivo and in vitro. However, this antithrombotic effect of magnesium may be beneficial in some patients after coronary revascularization, but magnesium therapy should be carefully considered in patients with preoperative ASA and antiplatelet drug use, impaired platelet function and coexisting bleeding disorders. Whether magnesium exerts an inhibitory effect on platelet function in patients receiving heparin and aspirin in OPCAB has not been investigated.13 Platelet dysfunction is the most important etiologic factor in the hemostatic defect observed following the heparin usage.18, 19, 20, 21 Kynczl-Leisure examined the effect of MS on the platelet function and fibrinolysis process when administered before the institution of CPB. Use of MS resulted in a significant prolongation of the BT, a significant reduction in the level of shed blood thromboxane B2, and an increase in the plasma levels of plasmin and D-dimer.22 Briel concluded that MS, independent of CPB, causes both platelet dysfunction, and increased fibrinolysis.3 Increasing postoperating bleeding, surgical re-exploration and transfusion volume reflect a significant statistical differences between two groups and in other hand, the percentage change of BT in group A was significantly longer than to group B. Mechanisms of postoperative bleeding in cardiac and noncardiac surgery are completely different.23 Low dose of MS (<2 g) without affecting BT with vasodilatation and reducing venous bleeding reduce total postoperative blood loss. But larger dose of MS in cardiac surgery not only causes tachycardia and venous dilatation but also leads to excessive postoperative bleeding with increasing BT.22 We found that hemodynamic changes of MS use are moderate hypotension and tachycardia that lasted for a short period and resolved spontaneously. Reducing blood loss in surgical field provides a better surgical field, in noncardiac surgery which is different in OPCAB surgery operations.23 Contrary to our results, Dabbagh et al. demonstrated that less postoperative bleeding and packed cell use in the MS group were observed in elective on-pump CABG surgery.24

5. Conclusions

These data indicate that however surgical and anesthesia times were not different between two groups, but postoperative diffuse bleeding was statistically different between two groups. This difference reveals a coagulopathy of MS that was not observed in group B. Platelet incorporation of MS at levels attained with preoperative IV supplementation is associated with inhibition of platelet aggregation through a PKC-dependent mechanism. These observations may represent one detrimental mechanism for the observed effect of MS in postoperative bleeding.

Conflicts of interest

The authors have none to declare.


1. Aglio L.S., Stamford G.G., Maddi R. Hypoma gnesaemia is common following cardiac surgery. J Cardiothorac Vasc Anesth. 1991;5:201–208. [PubMed]
2. Kaplan M., Kut M.S., Icer U.A. Intravenous magnesium sulfate prophylaxis for atrial fibrillation after coronary artery bypass surgery. J Thorac Cardiovasc Surg. 2003;125:344–352. [PubMed]
3. Briel R.C., Lippert T.H.H., Zahradnik H.P. Changes in blood coagulation, thrombocyte function and vascular prostacyclin synthesis induced by magnesium sulphate. Geburtshilfe Frauenheilkd. 1987;47:332–336. [PubMed]
4. Solomon A.J., Berger A.K., Trivedi K.K. The combination of propranolol and magnesium does not prevent postoperative atrial fibrillation. Ann Thorac Surg. 2000;69:126–129. [PubMed]
5. Najafi M., Hamidian R., Haghighat B. Magnesium infusion and postoperative atrial fibrillation: a randomized clinical trial. Acta Anaesthesiol Taiwan. 2007;45:89–94. [PubMed]
6. Hamid M., Kamal R.S., Sami S.A. Effect of single dose magnesium on arrhythmias in patients undergoing coronary artery bypass surgery. J Pak Med Assoc. 2008;58:22–27. [PubMed]
7. Cagli K., Ozeke O., Ergun K. Effect of low-dose amiodarone and magnesium combination on atrial fibrillation after coronary artery surgery. J Card Surg. 2006;21:458–464. [PubMed]
8. Zangrillo A., Landoni G., Sparicio D. Perioperative magnesium supplementation to prevent atrial fibrillation after off-pump coronary artery surgery: a randomized controlled study. J Cardiothorac Vasc Anesth. 2005;19:723–728. [PubMed]
9. Piper S.N., Kiessling A.H., Suttner S.W., Ducke M., Boldt J., Rohm K.D. Prevention of atrial fibrillation after coronary artery bypass graft surgery using a potassium–magnesium–aspartate solution (Inzolen) Thorac Cardiovasc Surg. 2007;55:418–423. [PubMed]
10. Maslow A.D., Regan M.M., Heindle S., Panzica P., Cohn W.E., Johnson R.G. Postoperative atrial tachyarrhythmias in patients undergoing coronary artery bypass graft surgery without cardiopulmonary bypass: a role for intraoperative magnesium supplementation. J Cardiothorac Vasc Anesth. 2000;14:524–530. [PubMed]
11. England M.R., Gordon G., Salem M. Magnesium administration and dysrhythmias after cardiac surgery. J Am Med Assoc. 1992;268:2395–2402. [PubMed]
12. Fanning W.J., Thomas C.S., Jr., Roach A. Prophylaxis of atrial fibrillation with magnesium sulphate after CABG. Ann Thorac Surg. 1991;52:529–533. [PubMed]
13. Gries A., Bode C., Gross S. The effect of intravenously administered magnesium on platelet function in patients after cardiac surgery. Anesth Analg. 1999;88:1213–1224. [PubMed]
14. Rawitscher R.E., Jones J.W., McCoy T.A. A prospective study of aspirin's effect on red blood cell loss in cardiac surgery. J Cardiovasc Surg. 1991;32:1–7. [PubMed]
15. Boldt J., Knothe C., Zickmann B. The effects of preoperative aspirin therapy on platelet function in cardiac surgery. Eur J Cardiothorac Surg. 1992;6:598–602. [PubMed]
16. Benedetto C., Massobrio M., Bertini E. Reduced serum inhibition of platelet-activating factor activity in preeclampsia. Am J Obstet Gynecol. 1989;160:100–104. [PubMed]
17. Caspi J., Rudis E., Bar I. Effects of magnesium on myocardial function after coronary artery bypass grafting. Ann Thorac Surg. 1995;59:942–947. [PubMed]
18. Harris J.E., Crowther A., Jupp R.A. Magnesium and coronary revascularization. Br J Anaesth. 1988;60:779–783. [PubMed]
19. Falck G., Lundgaard H., Jareld T. Effect of magnesium infusion on bleeding time in healthy male volunteers. Scand J Clin Lab Invest. 1999;59:425–430. [PubMed]
20. Mak T., Komarov A.M., Kramer J.H. Protective mechanisms of Mg-gluconate against oxidative endothelial cytotoxicity. Cell Mol Biol. 2000;46:1337–1344. [PubMed]
21. Jang E.K., Azzam J.E., Dickinson N.T. Roles for both cyclic GMP and cyclic AMP in the inhibition of collagen-induced platelet aggregation by nitroprusside. Br J Haematol. 2002;117:664–675. [PubMed]
22. Kynczl-Leisure M., Cibils L.A. Increased bleeding time after magnesium sulfate infusion. Am J Obstet Gynecol. 1996;175:1293–1294. [PubMed]
23. Rukshin V., Azarbal B., Shah P.K. Intravenous magnesium in experimental stent thrombosis in swine. Arterioscler Thromb Vasc Biol. 2001;21:1544–1549. [PubMed]
24. Dabbagh A., Rajaei S., Shamsolahrar M.H. The effect of intravenous magnesium sulfate on acute postoperative bleeding in elective coronary artery bypass surgery. J Perianesth Nurs. 2010;25:290–295. [PubMed]

Articles from Indian Heart Journal are provided here courtesy of Elsevier