Platelet Function Test in Coronary Artery Bypass Grafting: Does It Predict Postoperative Bleeding?

Introduction

Cardiac surgery using cardiopulmonary bypass (CPB) creates a risk of perioperative excessive bleeding; transfusion of red blood cells; transfusion of procoagulant blood products like fresh frozen plasma (FFP), platelet concentrate, and cryoprecipitate; and complications associated with transfusion.¹

From clinical studies, we have strong evidence that blood and blood products' transfusions during cardiac surgery have increased risk of serious postoperative morbidities including infections, atrial fibrillation, respiratory complications, acute kidney injury, duration of intensive care unit (ICU) and hospital stay, treatment costs, and short-term and long-term mortality.², ³

Excessive bleeding during and after cardiac surgery has an incidence of approximately 20%.⁴ Excessive bleeding and subsequent requirement for blood product administration and mediastinal reexploration is associated with significant morbidity and mortality.⁵ Reoperation for bleeding is required in 2.3 to 8% patients undergoing coronary artery bypass graft (CABG) surgery.⁶, ⁷, ⁸

Coagulopathy associated with cardiac surgery is multifactorial. Platelet dysfunction, excessive fibrinolysis, hypothermia, preoperative anemia, and deficiency of coagulation factors or their dilution are all suggested etiologies of postoperative bleeding.⁴

Among these, the most important is thought to be platelet dysfunction which occurs as a result of the interplay of acquired and pharmacologically induced factors.⁹, ¹⁰ Patients suffering from coronary artery disease are usually advised to stop antiplatelet medication a few days prior to CABG surgery to reduce the incidence of postoperative bleeding.¹¹ Current revascularization guidelines therefore recommend that clopidogrel and ticagrelor are discontinued 5 days before surgery and prasugrel 7 days before elective surgery, but the patient's condition may render this impossible.¹², ¹³ Patients who are still on antiplatelet drugs are at an increased risk of postoperative bleeding.¹⁴, ¹⁵ Transfusions of blood and blood products have got deleterious effects, and hence efforts should be made to minimize transfusion.¹⁶ Currently, the transfusion of blood and blood components to manage postoperative bleeding after CABG remains largely empirical, with considerable variation among institutions.¹⁷ Algorithm-based coagulation therapy has been shown to be superior to empiric therapy based on clinical judgment.⁸, ⁹, ¹⁰ Hence, there is a need to have objective tests to demonstrate platelet dysfunction before platelet transfusion.¹⁷, ¹⁸ Several devices' platelet function tests have been reported in clinical studies to evaluate platelet dysfunction and quantify the need for antiplatelet therapy.¹⁴, ¹⁷ Those devices range from highly sophisticated ones, such as light transmittance aggregometers and whole blood flow cytometers, to simple point-of-care technologies. The problem is that no consensus exists concerning several issues such as which device is optimal, what cut-off values to apply for discrimination of bleeding risk, and how to proceed with antiplatelet strategies once the results of platelet testing are determined.¹⁹

We hypothesized that objective measurement of platelet function can be useful in prediction of postoperative blood loss and in guiding postoperative transfusion requirements (blood and platelets) and the need for surgical reexploration. Hence, we designed a prospective observational study to investigate the association between platelet function by ROTEM platelet aggregometry and postoperative blood loss.

Materials and Methods

This prospective observational study was performed in the Department of Cardiac Anaesthesiology, All India Institute of Medical Sciences (AIIMS), New Delhi, after approval of the Institutional Ethical committee.

Approval of Institutional Ethical Committee was obtained vide letter number IECPG-583/24.10.2019, RT-12/28.11.2019 dated November 11, 2019. A written informed consent was obtained from all the patients before the screening in the study.

It is a prospective observational study and 52 patients who underwent on-pump CABG were analyzed with ROTEM platelet aggregometry in this investigation.

Methods of Collection of Data

Methodology

Preanesthetic examinations were performed 1 day prior to surgery. The procedure was explained to the patient and written informed consent were obtained. All patients received premedication in the form of injection morphine 0.1 mg/kg and injection promethazine 0.5 mg/kg. Anesthesia was induced with etomidate (0.2–0.3 mg/kg) or thiopentone (3–5 mg/kg), fentanyl (2–3 µg/kg) and rocuronium bromide (0.8–1 mg/kg). Patient was intubated with portex endotracheal tube of appropriate size. Anesthesia was maintained using sevoflurane (0.5–1%) or isoflurane (0.4–0.8%) in oxygen air mixture with intermittent doses of fentanyl, midazolam, and cisatracurium or vecuronium which is the standard practice in our institute. All patients were monitored with five-lead electrocardiography, SPO₂, invasive blood pressure, central venous pressure, temperature, capnography, and urine output.

Patients were operated according to standard protocol with CPB. Before cannulation, heparin (400 IU/kg) was given and supplemented as required to maintain an Activated Clotting Time (ACT) > 480 seconds. The standard uncoated CPB circuit was primed with 1-L lactated ringer solution, sodium bicarbonate of 50 mL, 0.5 gm/kg mannitol, and 10,000 IU heparin. The ascending aorta and atriocaval cannulation were done for CPB. CPB was performed with membrane oxygenator (CAPIOX FX25 with hard-shell reservoir) using standard nonpulsatile technique. After decannulation, heparin was neutralized with protamine sulfate (1-mg protamine per 100-IU heparin). If ACT was more than 170 seconds or baseline after heparin reversal, a supplementary dose of 50-mg protamine was given.

Patients were given packed red blood cell (PRBC), and blood components after following an AIIMS algorithm–based transfusion of blood components (Fig. 1). This algorithm will aid us to improve outcomes immensely, especially in postoperative bleeding control. It is a ROTEM-based platelet function tests using three different indicator agents, approved by our Ethics Committee.²⁰

Close

Fig. 1

Algorithm used at the All India Institute of Medical Sciences (AIIMS) for testing platelet aggregometry in coronary artery bypass grafting patients. A6, amplitude after 6 minutes; AUC, area under the curve; FFP, fresh frozen plasma; MS, maximum slope.

Export to PPT

PRBCs were added to the reservoir to maintain a hematocrit greater than 30%. ACT and blood gas analysis were done half hourly. Pump flow was maintained between 2.2 and 3 L/m² body surface area. The target mean arterial pressure during CPB was 60 mm Hg. If necessary, epinephrine was given to reach this value. Distal coronary anatomizes were performed on an arrested heart during a single period of aortic cross-clamping. Weaning from CPB was initiated once the patient's rhythm had stabilized and normothermia had been achieved. Isotropic support was initiated to maintain a cardiac index 2.2 L/min/m². The preferred isotropic agent was dobutamine. Nor epinephrine was used if dobutamine produced excessive vasodilatation. Epinephrine or dopamine was used if the hemodynamic performance remained inadequate. Postoperative bleeding was defined as the total amount of chest tube drainage during the first 24 hours. After surgery, the decision to transfuse PRBCs was based on clinical and hemodynamic status with aim of hemoglobin 10 gm/dL. FFP was transfused for ongoing significant bleeding and increased INR. Platelet was transfused if platelet count was <100 × 10⁶/dL and significant bleeding. Blood samples for ROTEM platelet aggregometry were obtained at two different time points: after induction of anesthesia before heparinization (T1), and 5 to 10 minutes after protamine administration (T2) in a tube-containing citrate and analyzed within 1 hour, after decalcification with 20-μL CaCl₂. All measurements were performed by physicians trained to operate study device and not directly involved in the patient's care.

Anesthesiologists and surgeons were blinded to the results of the ROTEM platelet measurements. Results of ROTEM platelet were not available to the operating room staff or the ICU team. Surgical reexploration of the mediastinum for excessive bleeding was noted, along with any surgical explanation of bleeding. An AIIMS algorithm–based approach was applied to transfuse according to the platelet reagent activates of ADPTEM, ARATEM, or TRAPTEM as depicted in Fig. 1.

Results

A total of 75 patients were assessed for eligibility. Four patients did not give consent for participation. Six patients were on low molecular weight heparin due to significant left main disease and they were not included due to not meeting inclusion criteria. Four patients had chest reexploration and due to an obvious surgical cause of bleeding, they were excluded for analysis. Six patients were excluded from data analysis due to a lack of reagents. Three patients had sudden hemodynamic instability postoperatively followed by intra-aortic balloon pump insertion prior, and they were excluded from study. The final sample size was 52.

Receiver Operating Characteristic Curve and Its Interpretation: At 5 to 10 Minutes after Protamine Administration for Amplitude after 6 Minutes

Interpretation

The area under the ROC curve at T2 for the parameter A6 is given above (Fig. 2). For TRAPTEM, the area is 50.2% (p = 0.982); for ARATEM, the area is 54.5% (p = 0.646); and for ADPTEM, the area is 67.4% (p = 0.079). None of these are significant and hence may not be good indicators for predicting excessive bleeding (Fig. 3).

Close

Fig. 2

Receiver operating characteristic (ROC) curve at T2 for A6. A6, amplitude after 6 minutes; T2, 5–10 minutes after protamine administration.

Export to PPT

Close

Fig. 3

Receiver operating characteristic (ROC) curve at T2 for MS. MS, maximum slope; T2, 5–10 minutes after protamine administration.

Export to PPT

Receiver Operating Characteristic Curve and Its Interpretation: At 5 to 10 Minutes after Protamine Administration for Maximum Slope

Interpretation

The area under the ROC curve at time T2 for the parameter MS is given above. For TRAPTEM, the area is 50.9% (p = 0.929); for ARATEM, the area is 48.6% (p = 0.884); and for ADPTEM, the area is 74.4% (p = 0.014). Only ADPTEM is found to be significant. For this, a threshold value of ≤3.5 to predict excessive bleeding yields sensitivity of 63.64%, specificity of 63.41%, PPV of 31.32%, NPV of 86.67%, and accuracy of 63.46% (Fig. 4).

Close

Fig. 4

Receiver operating characteristic (ROC) curve at T2 for AUC. AUC, area under the curve; T2, 5–10 minutes after protamine administration.

Export to PPT

Receiver Operating Characteristic Curve and Its Interpretation: At T2 for Area under the Curve

Interpretation

The area under the ROC curve at time T2 for the parameter AUC is given above. For TRAPTEM, the area is 70.3% (p = 0.040); for ARATEM, the area is 73.9% (p = 0.016); and for ADPTEM, the area is 78.5% (p = 0.004). For TRAPTEM, a threshold value of ≤50 to predict excessive bleeding yields sensitivity of 90.91%, specificity of 68.29%, PPV of 43.48%, NPV of 96.55%, and accuracy of 73.08%. For ARATEM, a threshold value of ≤31.5 to predict excessive bleeding yields sensitivity of 81.82%, specificity of 65.85%, PPV of 39.13%, NPV of 93.10%, and accuracy of 69.23%. For ADPTEM, a threshold value of ≤32 to predict excessive bleeding yields sensitivity of 72.73%, specificity of 87.80%, PPV of 61.54%, NPV of 92.31%, and accuracy of 84.62% (Fig. 5).

Close

Fig. 5

Receiver operating characteristic (ROC) curve for CPB time and Ao X clamp time. Ao X, aortic cross; CPB, cardiopulmonary bypass.

Export to PPT

Discussion

Antiplatelet therapy (APT) is the mainstay of treatment in patients with coronary artery disease. APT causes acquired platelet dysfunction and this group of patients are at high risk for excessive microvascular bleeding and reexploration after CABG. Postoperative bleeding remains one of the major complications of cardiac surgery, and the pivotal role of platelet function must be taken into consideration.²³ Effect of CPB itself and long CPB causes deranged coagulation and postbypass nonsurgical bleeding. Haensig et al concluded that in patients with long CPB times, ROTEM-guided treatment may result in less bleeding, a marked reduction in costs, and long-term mortality.²³

Despite this evidence base and the publication of numerous practice guidelines,²⁵ much confusion remains about the optimal management of perioperative bleeding in cardiac surgery patients. In contrast with other studies,²⁶ we focused on changes in platelet function and identifying the optimum time point for platelet function testing to predict bleeding and transfusion in this setting using ROTEM platelet.

We found significant difference in platelet function were present at T1 and T2 (TRAPTEM, ARATEM, and ADPTEM) in both excessive and nonexcessive bleeding groups. The most consistent results with significant differences in platelet function values between patients with and without excessive bleeding were found at T2 (5–10 minutes after protamine administration) in AUC (TRAPTEM, ARATEM, and ADPTEM). This is consistent with other studies demonstrating the inhibitory effect of CPB on platelet function assay. Velik-Salchner et al found that in patients without antiplatelet drugs, both multiplate and light transmission aggregometry (LTA) methods showed that CPB was associated with significantly decreased platelet aggregation after activation by commonly used agonists. This was also observed in patients taking aspirin or dual antiplatelet therapy for ADP-induced assays.²⁷

In our study results, we found no significant difference in preoperative platelet count, nor was the incidence of preoperative intake of antiplatelet drugs, such as aspirin and clopidogrel, different between patients with and without excessive bleeding. In contrast, the presence of excessive bleeding could be discriminated by platelet function testing at T2. Haas et al concluded that standard laboratory tests were neither designed to accurately reflect compromised thrombin generation nor to guide coagulation therapy and low predictive value of standard laboratory tests to predict postoperative bleeding.²⁸ We could not establish the history of discontinuation of antiplatelet therapy as a predictor of postoperative blood loss. We demonstrated that platelet function analysis is superior to history of discontinuation of antiplatelet therapy for predicting bleeding in cardiac surgery. Rahe-Meyer et al also suggested that arachidonic acid-induced aggregation in whole blood may be a better predictor of platelet-related coagulopathy and platelet transfusion than the assessment of aspirin intake by patient self-reporting.²⁹

For predicting the bleeding of ROC curve was evaluated for each parameter of all three indices of ROTEM platelet at T2 point. For MS area under the ROC curve, only ADPTEM was found to be significant. For this, a threshold value of ≤3.5 to predict excessive bleeding yields sensitivity of 63.64%, specificity of 63.41%, PPV of 31.32%, NPV of 86.67%, and accuracy of 63.46%. Both sensitivity and specificity were low with NPV of 86.67%. For AUC, for TRAPTEM, a threshold value of ≤50 to predict excessive bleeding yielded sensitivity of 90.91%, specificity of 68.29%, PPV of 43.48%, NPV of 96.55%, and accuracy of 73.08%. For ARATEM, a threshold value of ≤31.5 to predict excessive bleeding yielded sensitivity of 81.82%, specificity of 65.85%, PPV of 39.13%, NPV of 93.10%, and accuracy of 69.23%. For ADPTEM, a threshold value of ≤32 to predict excessive bleeding yielded sensitivity of 72.73%, specificity of 87.80%, PPV of 61.54%, NPV of 92.31%, and accuracy of 84.62%.

Cut-off values, as determined by ROC analyses, had a consistently weak PPV for all tests at T2 time point, whereas NPVs were higher and generally had an increasing tendency toward measurements performed at T2. Ranucci et al⁵ evaluated prediction bleeding using whole blood impedance aggregometry, and reported a low PPV coupled with a high NPV. This was superior in estimating who was likely to not bleed rather than who was likely to bleed. They pointed out that the high NPV of whole blood impedance aggregometry helps to exclude platelet dysfunction as the reason for bleeding after cardiac surgery in case of normal results in the activated by adenosine diphosphate (ADP) activated assay (negative predictive value of 92%).

Petricevic et al assessed platelet function by two whole blood impedance aggregometers (ROTEM platelet and Multiplate, using three different activators (arachidonic acid, adenosine diphosphate, and thrombin receptor-activating peptide-6), at three perioperative time points (before anesthesia, after aortic declamping, and 5–10 minutes after protamine administration).³⁰ Multiplate ADP test AUC cut-off value was ≤27 U to delineate excessive bleeding. For the ROTEM platelet assay ADPTEM, the corresponding AUC cut-off value was ≤36 Ohm.min. For TRAP test (multiplate) and TRAPTEM (ROTEM platelet), the AUC cut-off value for excessive bleeding was ≤77 U and ≤46 minutes, respectively.

Bleeding during and after cardiac surgery usually is multifactorial and platelet function analysis cannot rule out surgical bleeding, as well as coagulopathic bleeding due to hyperfibrinolysis, low fibrinogen, or impaired thrombin. There are many studies demonstrating the predictive value of point-of-care platelet function testing using whole blood impedance aggregometry for postoperative blood loss and transfusion requirements in cardiac surgery.³¹, ³² In our study, we assessed prediction of bleeding using ROTEM platelet aggregometry at two different points. At T1, we assessed difference in platelet function due to preoperative factors, like antiplatelet therapy, and at T2, we evaluated the effect of CPB on platelet function and analyzed their predictive values for postoperative blood loss and gave blood according to an algorithm as depicted in Fig. 1.

We found that platelet function analysis using multiple (Activated by arachidonic acid (ARA), ADP, and Activated by thrombin receptor-activating peptide-6 (TRAP)) activators after CPB and protamine administration is able to predict postoperative blood loss with high NPV and low NPV. It helps to exclude platelet dysfunction as the reason for bleeding after cardiac surgery. Point-of-care platelet function analysis, particularly in combination with viscoelastic testing, such asthromboelastometry (ROTEM), can reduce perioperative bleeding and transfusion requirements, as well as improve patient outcomes in cardiac surgery. Earlier too, using algorithm and Point of care (POC) testing, the authors have controlled bleeding using the right blood component therapy but in CABG patients, the postoperative transfusion in the excessive bleeding group helped in optimizing patient control. Platelet function test (PFT)-based Point of care (POC) used in an algorithmic manner is where on-pump CABG bleeding control lies today.

Limitations of Study

In cardiac surgery, there is no agreement on the definition of excessive bleeding. Chest tube drainage does not reflect intraoperative blood loss and transfusion requirements and consists of a mixture of fluids, including actual blood loss, serous drainage, and fluid left in the pleural cavity. It may, therefore, overestimate postoperative blood loss. Our sample size is small because of poor availability of regent and affordability of this costly test. We did not have any provision to compare our test results with established testing platforms like mutilate. Furthermore, the pharmacokinetic analysis of drug levels in blood was not performed in the study.

There are many factors that can influence blood loss after cardiac surgery, we cannot expect strong predictions from a bed-side test assessing only one part of hemostasis, that is, platelet aggregation. Although, there is consensus regarding the dominant role of platelets in hemostasis after cardiac surgery, we believe that adequate interaction of all coagulation components is crucial to achieve adequate hemostasis. For this reason, concomitant use of the platelet aggregometry and thromboelastometry, both pre- and intraoperatively, could provide a complete and more reliable picture of hemostatic disturbances. This would also provide the role of fibrinogen and its interaction with platelets.

In our study, the daily dose of antiplatelet medicine received were not documented. In our study patients undergone CABG had different anesthetic and perfusion teams. Surgery was performed in three units. But all patients had same anesthesia protocol and same standard surgical procedure. In our study, chest drain output was not adjusted to body weight. But in final analysis, we compared the chest drain output with body weight and body mass index.

Conclusion

Platelet function analysis using multiple (ARA, ADP, and TRAP) activator after protamine administration is able to predict postoperative blood loss with high NPV and low PPV. Therefore, it helps to exclude platelet dysfunction as the reason for bleeding after cardiac surgery. Point-of-care platelet function analysis, particularly in combination with viscoelastic testing such as thromboelastometry (ROTEM), can reduce perioperative bleeding and transfusion requirements, as well as improve patient outcomes in cardiac surgery. Incorporation of platelet function test in transfusion algorithm may further improve patient outcome post cardiac surgery.