Asian Cardiovasc Thorac Ann 2006;14:310-315
© 2006 Asia Publishing EXchange Ltd
Using Cone and Plate(let) Analyzer to Predict Bleeding in Cardiac Surgery
Rabin Gerrah, MD,
Alex Brill, PhD1,
Sagi Tshori, MD2,
Aharon Lubetsky, MD3,
Gideon Merin, MD4,
David Varon, MD1
Department of Cardiothoracic Surgery, Assuta Medical Center, Tel Aviv, Israel
1 Coagulation Unit, Department of Hematology
2 Department of Medical Biochemistry, Hebrew University, Hadassah Medical School, Jerusalem, Israel
3 Institute of Thrombosis and Hemostasis, Sheba Medical Center, Tel Hashomer, Israel
4 Department of Cardiothoracic Surgery, Hebrew University, Hadassah Medical School, Jerusalem, Israel
For reprint information contact: Rabin Gerrah, MD Tel: 972 3 520 1436 Fax: 972 3 520 1747 Email: rabin{at}assuta.com, Cardiothoracic Surgery, Assuta Medical Center, 62 Jabotinsky Street, Tel Aviv 62748, Israel.
 |
ABSTRACT
|
|---|
The cone and plate(let) analyzer is an established method for assessing platelet function. It evaluates adherence of platelets on an extracellular matrix, expressed as a percentage of surface coverage and the average size of the aggregates. The purpose of this study was to determine the applicability of the cone and plate(let) analyzer in monitoring platelet function and predicting postoperative bleeding. The relationship between postoperative bleeding, perioperative platelet function, and other parameters was studied. A significant decrease in surface coverage was detected upon establishment of cardiopulmonary bypass (from 6.9% ± 3.9% to 4.7% ± 1.7%) with a return to preoperative values at the end of surgery. Preoperative average size and surface coverage were the only parameters that significantly and linearly correlated with postoperative bleeding. Patients with an aggregate average size < 20 µm2 had a significantly higher incidence of severe bleeding (> 965 mL) than those with a size > 20 µm2 (44% vs. 0%), and a higher mean blood loss (908 ± 322 mL vs. 337 ± 78 mL). Similar results were obtained for surface coverage < 5%, indicating the predictive value of these parameters. Preoperative platelet function as evaluated by the cone and plate(let) analyzer is an independent risk factor determining postoperative bleeding.
|
INTRODUCTION
|
|---|
Platelet dysfunction is common during cardiac surgery.1,2 Cardiopulmonary bypass (CPB) and other nonphysiologic factors impair platelet function and cause coagulation disorders and endothelial injury.1,3 Along with dysfunction, activation of platelets during CPB releases multiple mediators with a significant and broad spectrum of systemic effects.4 In addition, a considerable number of patients are treated with antiplatelet agents preoperatively. The main result of platelet dysfunction is a bleeding diathesis. In a few cases, a thrombotic complication can result. Perioperative bleeding has an important impact on prognosis, hence the need for a reliable method of testing platelet function. Various techniques have been described for evaluation of platelet function. The disadvantages of most of these are nonphysiologic conditions, complexity, time consumption, and need for special blood processing procedures. The cone and plate(let) analyzer (CPA) was designed to test platelet function under near-physiologic conditions.5,6 The test rapidly identifies both congenital and acquired platelet defects and the effect of antiplatelet drugs, and detects prothrombotic states with platelet hyperfunction.7–9 The CPA has proved useful in assessment of platelet function in various clinical conditions including thrombocytopenia after chemotherapy and monitoring treatment with antiplatelet drugs.10–12 The CPA yields two parameters: average size and surface coverage, which determine platelet function in terms of adhesion and aggregation. These values constitute a general platelet function parameter. Determination of a specific defect in platelet function might require more sophisticated tests. However, the identification of platelet dysfunction during heart surgery usually requires a platelet transfusion. In this study, we evaluated the applicability of the CPA test in monitoring platelet function during cardiac surgery.
|
PATIENTS AND METHODS
|
|---|
The study population consisted of 32 consecutive patients undergoing first-time cardiac surgery with CPB. Their procedures included coronary artery bypass grafting (CABG), valve surgery (repair or replacement), and combined operations (CABG and valve procedure). The institutional Human Ethics Committee was consulted on the need for approval. As all results were taken from routine tests, and no invasive procedures were carried out for the purpose of the study, the need for approval was waived. All patients had the same standard protocol of anesthesia and perioperative care. Heparin was injected at 3 mg·kg–1 and anticoagulation was monitored to achieve an activated clotting time < 480 sec. Additional heparin was injected if needed. The membrane oxygenator and the pump tubes (Cobe Cardiovascular, Inc., Arvada, CO, USA) were identical in all patients. Heparin-coated tubes were not used. Patients with a known coagulation disorder or an abnormality of platelet function or count and those treated with antiplatelet agents such as glycoprotein IIb/IIIa antagonists were excluded from the study. Demographic data, all perioperative factors with a potential effect on postoperative bleeding, and medications with a possible effect on platelet function or count were recorded (Table 1
). Blood samples were obtained from those drawn for routine monitoring during surgery: during induction of anesthesia; immediately after mid sternotomy; 10 min after heparin injection; 10 and 30 min after establishment of CPB; after discontinuation of CPB; 10 min after protamine injection; and at the end of the operation. At each time point, a 5 mL sample of blood was stored in a citrated tube until the end of the operation. A complete blood count using an automatic counter and evaluation of platelet function by the CPA were performed simultaneously.
For CPA evaluation, 200 µL of whole blood in 3.8% sodium citrate solution was placed in a polystyrene plate (Nunc Multidish 4UBEH, Nalge Nunc international, Roskilde, Denmark), and a defined shear rate (1,800 sec–1) was applied for 2 minutes using a rotating Teflon cone. This was followed by washing off non-adherent blood elements, staining, and measuring the plate surface coverage and average size of the adherent aggregates of platelets, using an image analyzer. Under these test conditions, only platelets adhere to the plate surface. When normal blood is analyzed, platelet deposition is a shear- and time-dependent process, reaching maximal levels within 2 min at a high shear rate (1,800 sec–1), with approximately 15% surface coverage and 25–35 µm2 average size. The activity of platelets during hemostasis is determined by their adhesion and aggregation. Average size determines the mean size of the platelet aggregates and surface coverage measures the percentage of the surface covered by the platelet aggregates. Thus, average size and surface coverage are measures of the platelet functions of aggregation and adhesion. Normal values for average size and surface coverage in the cardiovascular population are 39 ± 5 µm2 and 15% ± 2.5%, respectively.
The indications for blood transfusion were active bleeding or a hemoglobin level < 8 mg·dL–1 in a stable patient or < 10 mg·dL–1 in an unstable patient or those symptomatic for anemia. During the operation, blood was added to the extracorporeal system to maintain a hematocrit of approximately 25%. The indication for plasma transfusion was a coagulation disorder in a bleeding patient. Transfusions of platelets, fresh frozen plasma, and cryoprecipitate were used in patients with active postoperative bleeding when it was due to a coagulation defect rather than a surgical cause.
Postoperative care was standard for all patients according to a predetermined protocol. All parameters related to bleeding or thrombotic events were recorded. Postoperative bleeding was measured as the total amount of blood in the collection system from the time of chest closure until removal of the chest tubes. Chest tubes were withdrawn 24 hours postoperatively in all but 2 patients who had continuing discharge; their chest tubes were removed after 48 hours. The total units of blood or blood product transfused, exceptional bleeding, and the need for re-exploration for pericardial tamponade or major bleeding were recorded. Data were analyzed at both the univariate and multivariate levels. Data were entered into a statistical database as either continuous or categorical variables for comparative statistical analysis. Data are expressed as mean ± standard deviation and range, where applicable. The univariate analyses included Student’s t test for the comparison of two groups of quantitative variables, the chi-squared test for assessment of the association between two categorical variables, and the Pearson correlation coefficient for calculating the correlation between two quantitative variables. The influence of different variables on postoperative bleeding was evaluated by a multivariate general linear model analysis. A p-value of 0.05 or less was considered significant.
|
RESULTS
|
|---|
The preoperative and operative data are summarized in Table 1
. There were no differences in patient demographics or preoperative data among the different types of surgery. There was no perioperative mortality. The CPA results showed that the aggregate average size decreased to its lowest value of 22.4 ± 5.4 µm2 during CPB; however, this was not statistically significant. Furthermore, the average size at the end of the operation was not different from the preoperative value; p = 0.1 (Figure 1). Surface coverage at the start of surgery was 6.9% ± 3.9%. It increased initially and then declined to 4.7% ± 1.7% on institution of CPB, significantly lower than the preoperative value of 6.9% ± 9%; p = 0.008. The surface coverage increased in the latter stages of the operation after discontinuation of CPB, however, at the end of the operation it was similar to the preoperative level; p = 0.2 (Figure 2). The mean platelet count before the operation was 177 ± 65.5 x 1,000·µL–1. It decreased significantly, especially upon initiation of CPB and fell to a nadir of 109 ± 40 x 1,000·µL–1 ( p = 0.00001) after 30 min of CPB. It recovered towards the end of the operation, but the final count (141.3 ± 51.7 x 1,000·µL–1) was still significantly lower than the preoperative count ( p = 0.02).
View larger version (13K):
[in this window]
[in a new window]
|
Figure 1. Changes in aggregate average size during cardiac surgery. CPB = cardiopulmonary bypass, Postop = Postoperative, Preop = Preoperative.
|
|
View larger version (13K):
[in this window]
[in a new window]
|
Figure 2. Changes in platelet surface coverage during cardiac surgery. CPB = cardiopulmonary bypass, Postop = Postoperative, Preop = Preoperative.
|
|
The relationships between CPB time and postoperative bleeding, hematocrit, platelet count and function were determined to assess the impact of CPB. No linear relationship was found between duration of CPB and postoperative bleeding. Postoperative platelet function, measured as average size and surface coverage, was not influenced by the duration of CPB; r = –0.4, p > 0.05. However, the CPB time correlated weakly with the postoperative platelet count; r = –0.4, p = 0.05. A trend towards an association was noted between postoperative hematocrit and CPB time; r = –0.3, p = 0.06. The mean postoperative blood loss was 630 ± 380 mL. One patient underwent re-exploration because of uncontrollable bleeding and hemodynamic instability with suspected tamponade. In univariate analysis, the association of postoperative bleeding with several perioperative factors with a potential effect on bleeding was evaluated. A significant negative correlation was found between postoperative bleeding and preoperative platelet function measured by the CPA, both as average size (r = –0.61, p = 0.003) and surface coverage (r = –0.62, p = 0.001). The correlation of postoperative bleeding with platelet function by CPA was not significant for postoperative average size, and borderline for postoperative surface coverage (r = –0.4, p = 0.08). Interestingly, platelet count had no significant correlation with postoperative bleeding at any surgical stage. The association between the lowest platelet count during CPB and postoperative bleeding was examined to determine whether the severity of this decline could be predictive of bleeding. This association was not statistically significant. The type of surgery also had no effect on the CPA parameters or postoperative bleeding. Univariate analysis revealed no significant correlation, except for preoperative platelet function (average size and surface coverage), between the postoperative blood loss and any perioperative factor incorporated into the study as a variable. Two of the 32 patients had taken aspirin until the operation. The specific data on platelet function by CPA in this subgroup are shown in Figure 3. The total blood loss in these two patients was 555 mL and 1,095 mL.
View larger version (15K):
[in this window]
[in a new window]
|
Figure 3. Changes in average size and surface coverage during sugery in 2 patients on aspirin. CPB = cardiopulmonary bypass, Postop = Postoperative, Preop = Preoperative.
|
|
The relationship between all parameters assessed by univariate analysis was evaluated in the multivariate analysis model. These parameters were always initially entered into the model as independent variables (some were considered to be confounders and others to be risk factors). Postoperative bleeding was found to be significantly and linearly dependent on the preoperative platelet function measured by average size and surface coverage ( p = 0.003 and 0.001, respectively, r = 0.6). None of the other independent variables were found to be significant. Postoperative bleeding was stratified into 3 subgroups; minor blood loss was defined as < 370 mL, moderate blood loss (371–965 mL), and major blood loss (> 965 mL). The correlation between these subgroups and the CPA tests revealed a significant negative correlation with preoperative average size. Patients with major postoperative bleeding had the lowest preoperative aggregate average size, and those with minor bleeding had the highest preoperative average size; similar findings were noted with preoperative surface coverage (Table 2). These subgroups had similar demographics and their preoperative data were not different regarding bleeding-related factors.
Patients were categorized by CPA average size into two groups: patients with aggregate average size < 20 µm2 had a higher incidence of severe bleeding than those with average size > 20 µm2 (44% vs. 0%), and a much higher blood loss (908 ± 322 vs. 337 ± 78 mL, p = 0.0006; Figure 4). Patients were also categorized by surface coverage into two groups: those with surface coverage < 5% had a higher incidence of severe bleeding than those with surface coverage > 5% (33% vs. 0%), and a higher blood loss (816 ± 346 vs. 352 ± 103 mL, p = 0.0006; Figure 4). Thus, it seems that both CPA average size and surface coverage can be used to predict the risk of severe bleeding and to assess expected blood loss. Blood and blood product transfusions included packed red cells, whole blood, fresh frozen plasma, platelets, and cryoprecipitate. Of the 32 study patients, 24 (75%) had perioperative blood transfusions. Fresh frozen plasma was transfused in 6 patients, platelets in 2, and cryoprecipitate in one. No difference was found between the percentage of patients who received transfusions and different subtypes of blood loss or preoperative CPA parameters (Table 2). The perioperative CPA platelet function studies revealed no case of prothrombotic levels of average size or surface coverage. In the early hospital follow-up period, no thromboembolic event was recorded in any group. In one patient, leg ischemia developed after the surgery, which was due to direct mechanical injury at the femoral artery puncture site.
View larger version (18K):
[in this window]
[in a new window]
|
Figure 4. Incidence of minor (< 370mL), moderate (371–965 mL), and major blood loss (> 965 mL) as a function of (a) average size (AS) and (b) surface coverage (SC). (c) Mean blood loss (mL) as a function of average size and surface coverage.
|
|
|
DISCUSSION
|
|---|
The significant decrease in platelet count observed following establishment of CPB, and the deterioration of platelet function during CPB might be explained by the effects of heparin and the extracorporeal circulation on blood components. Activation of platelets following heparin administration has been described previously.13 Vast exposure of platelets to the synthetic surfaces of tubes and the oxygenator causes massive activation and degranulation of platelets, resulting in deterioration of platelet function postoperatively, in terms of both aggregation and adhesion.14,15 Recovery of platelet function following weaning from CPB has been noted in this and other studies.16
Kunitomo and colleagues17 defined the factors determining postoperative platelet count: preoperative platelet count, age, and blood transfusions during the operation. The predictive factors for postoperative blood loss were age and amount of bleeding during surgery. Our finding of a significant correlation between blood loss and preoperative platelet function is important in postoperative management of patients in the intensive care unit. Patients with preoperative aggregate average size < 20 µm2 or surface coverage < 5% had a high postoperative blood loss (> 965 mL), whereas those with average size > 20 µm2 or surface coverage > 5% lost approximately 300 mL of blood after surgery. Thus, patients with average size < 20 µm2 or surface coverage < 5% would be at risk of bleeding complications, including viral infections as a result of transfusions, cardiac tamponade, and decreased graft flow.
The current widespread use of novel antiplatelet agents such as glycoprotein IIb/IIIa antagonists, and the increasing number of patients undergoing angioplasty and stenting who require such medication, is a challenge for the surgical and intensive care teams.18–20 Thus, the prediction of platelet function and postoperative blood loss is becoming more important. Use of the CPA for real-time measurement of platelet function could be important in the care of these patients. The main damage to platelet function occurs during the initial stage of CPB. No association was found between the duration of CPB and postoperative platelet function. Moreover, postoperative platelet function in 6 patients who had an especially long duration of CPB did not differ significantly from the others. It is possible that no difference was noticed because of the small study population. Alternatively, the CPA may be less sensitive to mechanical damage to platelets than to antiplatelet drugs. The major limitations of this study are the small number and heterogeneous characteristics of the study population, and the lack of correlation of data obtained by the CPA with other established measures of platelet reactivity. Additional studies with larger populations are needed to confirm these results. Nevertheless, our findings suggest an important relationship between preoperative platelet function and postoperative bleeding. The CPA method seems to be a useful tool for testing perioperative platelet function and may help in predicting postoperative blood loss.
|
REFERENCES
|
|---|
- Weerasinghe A, Taylor KM The platelet in cardiopulmonary bypass. Ann Thorac Surg 1998;66:2145–52.[Abstract/Free Full Text]
- Zilla P, Fasol R, Deutsch M, Laufer G, Wollenek G, Muller MM, et al. Whole blood aggregometry and platelet adenine nucleotides during cardiac surgery. Scand J Thorac Cardiovasc Surg 1988;22:165–9.[Medline]
- Wan S, LeClerc JL, Vincent JL. Inflammatory response to cardiopulmonary bypass: mechanisms involved and possible therapeutic strategies. Chest 1997;112:676–92.[Abstract/Free Full Text]
- Collman RW. Platelet and neutrophil activation in cardiopulmonary bypass. Ann Thorac Surg 1990;49:32–4.[Abstract]
- Varon D, Dardik R, Shenkman S, Kotev-Emeth S, Farzame N, Tamarin I, et al. A new method for quantitative analysis of whole blood platelet interaction with extracellular matrix under flow conditions. Thromb Res 1997;85:283–94.[Medline]
- Shenkman B, Savion N, Dardik R, Tamarin I, Varon D. Testing of platelet deposition on polystyrene surface under flow conditions by the cone and plate(let) analyzer: role of platelet activation, fibrinogen and von Willebrand factor. Thromb Res 2000;99:353–61.[Medline]
- Knobler H, Savion N, Shenkman B, Kotev-Emeth S, Varon D. Shear-induced platelet adhesion and aggregation on subendothelium are increased in diabetic patients. Thromb Res 1998;90:181–90.[Medline]
- Shenkman B, Linder N, Savion N, Tamarin I, Dardik R, Kennet G, et al. Increased neonatal platelet deposition on subendothelium under flow conditions: the role of plasma von Willebrand factor. Pediatr Res 1999;45:270–5.[Medline]
- Wang X, Dorsam RT, Lauver A, Wang H, Barbera FA, Gibbs S, et al. Comparative analysis of various platelet glycoprotein IIb/IIIa antagonists on shear-induced platelet activation and adhesion. J Pharmacol Exp Ther 2002;303:1114–20.[Abstract/Free Full Text]
- Kenet G, Lubetsky A, Shenkman B, Tamarin I, Dardik R, Rechavi G, et al. Cone and platelet analyser (CPA): a new test for the prediction of bleeding among thrombocytopenic patients. Br J Haematol 1998;101:255–9.[Medline]
- Varon D, Lashevski I, Brenner B, Beyar R, Lanir N, Tamarin I, et al. Cone and plate(let) analyzer: monitoring glycoprotein IIb/IIIa antagonists and von Willebrand disease replacement therapy by testing platelet deposition under flow conditions. Am Heart J 1998;135(5 Pt 2 Su):S187–93.[Medline]
- Shenkman B, Schneiderman J, Tamarin I, Kotev-Emeth S, Savion N, Varon D. Testing the effect of GPIIb-IIIa antagonist in patients undergoing carotid stenting: correlation between standard aggregometry, flow cytometry and the cone and plate(let) analyzer (CPA) methods. Thromb Res 2001;102:311–7.[Medline]
- Chiba Y, Morioka K, Muraoka R, Ihaya A, Kimura T, Uesaka T, et al. Effects of depletion of leukocytes and platelets on cardiac dysfunction after cardiopulmonary bypass. Ann Thorac Surg 1998;65:107–14.[Abstract/Free Full Text]
- Wahba A, Black G, Koksch M, Rothe G, Preuner J, Schmitz G, et al. Cardiopulmonary bypass leads to a preferential loss of activated platelets. A flow cytometric assay of platelet surface antigens. Eur J Cardiothorac Surg 1996;10:768–73.[Abstract]
- Hyde JA, Chinn JA, Graham TR. Platelets and cardiopulmonary bypass. Perfusion 1998;13:389–407.[Free Full Text]
- Kvarstein B, Vaage S, Tjonneland S, Cappelen C Jr, Aune H. The adhesiveness of blood platelets before, during and after cardiopulmonary bypass. Scand J Thorac Cardiovasc Surg 1977;11:57–9.[Medline]
- Kunitomo R, Tsurusaki S, Suzuki R, Takaji K, Moriyama S, Hagio K, et al. Predictive factors for platelet number after cardiopulmonary bypass and postoperative blood loss. ASAIO J 2002;48:671–4.[Medline]
- Levy JH, Smith PK. Platelet inhibitors and cardiac surgery. Ann Thorac Surg 2000;70(2 Suppl):S1–2.[Free Full Text]
- Despotis GJ, Gravlee G, Filos K, Levy J. Anticoagulation monitoring during cardiac surgery: a review of current and emerging techniques. Anesthesiology 1999;91:1122–51.[Medline]
- Coller BS. Monitoring platelet GP IIb/IIIa antagonist therapy. Circulation 1998;97:5–9.[Free Full Text]
TOP
ABSTRACT
INTRODUCTION
