|
|
||||||||
a Department of Cardiothoracic Surgery, Helsinki University Central Hospital, Helsinki, Finland
b Department of Anesthesiology and Intensive Care Medicine, Helsinki University Central Hospital, Helsinki, Finland
c Department of Hematology, Helsinki University Central Hospital, Helsinki, Finland
d Laboratory Division (HUSLAB), Helsinki University Central Hospital, Helsinki, Finland
e Department of Pediatrics, Jorvi Hospital, Helsinki University Central Hospital, Espoo, Finland
Accepted for publication September 24, 2007.
* Address correspondence to Dr Raivio, Department of Cardiothoracic Surgery, Helsinki University Central Hospital, Post box 340, Helsinki, FIN-00029 HUS, Finland (Email: peter.raivio{at}hus.fi).
| Abstract |
|---|
|
|
|---|
Methods: Comprehensive thrombophilia screening was performed in patients (n = 100) before elective coronary artery bypass grafting (CABG) with CPB. Markers of thrombin generation (prothrombin fragment F1+2), its procoagulant activity (soluble fibrin complexes), and a marker of fibrin degradation (D-dimer) were measured serially at eight time points before, during, and after CABG.
Results: Abnormal thrombophilia screening was common (44%). While patients with thrombophilic variables had higher preoperative prothrombin fragment F1+2 than others (median [interquartile range] 0.55 [0.34] vs 0.45 (0.21) nmol/L, p = 0.009) they did not have higher F1+2, D-dimer, or soluble fibrin complex levels during CPB or postoperatively than patients without thrombophilic variables.
Conclusions: Preoperative thrombophilic variables do not associate with perioperative thrombin generation or its procoagulant activity in patients undergoing CABG. Our results do not support routine thrombophilia screening before CABG.
| Introduction |
|---|
|
|
|---|
Several thrombophilic factors, including prothrombin G20210A gene mutation [6], APC resistance [7], and deficiencies of protein C [8, 9], protein S [7, 9], and antithrombin [10] have been shown to associate with increased basal thrombin generation. On the other hand, cardiopulmonary bypass (CPB) causes a progressive increase in thrombin generation and its procoagulant activity [11], which is further distinctly propagated by reperfusion after myocardial ischemia [12–14]. It is therefore plausible that thrombophilic factors might enhance the CPB-related activation of coagulation. A possible thrombophilia-induced increase in thrombin generation might have clinical consequences as thrombin generation during reperfusion after CABG was previously shown to associate with postoperative myocardial damage and increased postoperative pulmonary vascular resistance [14]. Moreover, levels of a marker of thrombin generation, prothrombin fragment F1+2 (F1+2), were shown to correlate with measurements of organ dysfunction after CABG, including left ventricular stroke work index, arterial partial pressure of oxygen (PaO 2)/fraction of inspired oxygen ratio, and serum creatinine [15]. Also, activated protein C, a natural anticoagulant activated by thrombomodulin-bound thrombin, was shown to associate dynamically with postoperative hemodynamic recovery after CABG [16].
The objective of this study was to test whether preoperative thrombophilic factors associate with increased generation of thrombin or its procoagulant activity during and after CPB. Because of the multiple interactions of thrombin we selectively analyzed markers of its generation prothrombin fragment F1+2 (F1+2), procoagulant activity (soluble fibrin complexes [SFC]), and a marker of fibrin turnover (D-dimer, a surrogate marker for both thrombin and plasmin function).
| Patients and Methods |
|---|
|
|
|---|
|
Blood Samples
Blood samples were collected at eight time points (A to H): immediately preoperatively (A), at 15 minutes of CPB (B), immediately before the release of the aortic clamp (C), 15 minutes after the release of the aortic clamp (D), 30 minutes (E) and six hours after protamine administration (F), and on the first (G) and fifth (H) postoperative days. Samples A through G were collected through a radial artery catheter and sample H was collected either through an atraumatic venipuncture or through a radial artery catheter. The radial artery line was flushed only with physiologic saline without heparin. The first 5 mL of each sample were discarded. Nine volumes of blood were collected on one volume of sodium citrate into vacuum test tubes. For measurements of F1+2, SFC, and D-dimer, the samples were cooled on ice and centrifuged (1,500 g for 10 minutes) at +4°C. Plasma was separated and stored at –80°C. For other coagulation measurements the samples were centrifuged (2,000 g for 10 minutes) at +20°C before separation of plasma. Homocysteine was measured from ethylenediaminetetraacetic acid plasma.
Laboratory Analyses
F1+2 was analyzed with an enzyme-linked immunoassay (Enzygnost F1+2micro, Dade Behring, Marburg, Germany). SFC and D-dimer were analyzed with immunoturbidimetric assays (STA-Liatest FM; Diagnostica Stago, Asnieres, France and Tina-quant D-Dimer; Roche Diagnostics, Mannheim, Germany). F1+2 and D-dimer were measured from all patients at all time points (see above) and SFC were measured from patients 1 to 80 at time points A to F.
The preoperative thrombophilia screen included the following analyses: Antithrombin and protein C activities were determined with chromogenic assays (Berichrom Antithrombin III [A], Berichrom Protein C, Dade Behring, Marburg, Germany). Free protein S was determined with an automated latex ligand immunoassay (Instrumentation Laboratory, Lexington, MA). Factor VIII was analyzed with a coagulometric method (Dade Behring). APC resistance was determined with a functional activated partial thromboplastin time (APTT)-based test with predilution of patient plasma with factor V-depleted plasma (Coatest APC Resistance V, Chromogenix, Lexington, MA). Factor V Leiden (FVR506Q) and the prothrombin G20210A gene mutations were identified with cyclic minisequencing. Cardiolipin immunoglobulin G (IgG) antibodies were determined with an enzyme immunoassay (Pharmacia Deutschland, Freiburg, Germany). Lupus anticoagulant was screened with a simplified dilute Russell viper venom time test and confirmed with a high phospholipid-containing coagulation assay (DVVtest; DVVconfirm, American Diagnostica, Stamford, CT). Homocysteine was determined with a fluorescence polarization immunoassay (AxSYM Homocysteine, Axis-Shield, Dundee, UK).
Thrombophilic factors were defined as the following: presence of factor V Leiden, abnormal functional APC resistance test (APTT ratio <2.0) to demonstrate possible acquired APC resistance; prothrombin gene mutation G20210A; anticardiolipin IgG antibodies; lupus anticoagulant; antithrombin level below 67%; protein C or protein S level below the reference range; or factor VIII or homocysteine level above the reference range (Table 2).
|
The data were analyzed with the SPSS 11.5.1 software (SPSS, Chicago, IL) and the NCSS 2000 software (NCSS, Kaysville, UT). Normality of the distributions of the continuous variables was analyzed with the Kolmogorov-Smirnov test. Data are presented as median and interquartile range or mean and standard deviation, as appropriate. For clarity, data in figures are expressed as mean and standard error of the mean. Differences between the repeated measurements were analyzed with repeated measures analysis of variance (ANOVA) and post hoc comparisons were made with the Fisher least significant difference multiple comparison test. Variables with skewed distributions were natural logarithmically transformed before these analyses. The Mann-Whitney U test was used for comparisons between two groups and associations between two dichotomous variables were tested with the
2 test. For multivariate logistic regression analysis a forward stepwise method was used. Relevant patient characteristics (age, sex, and the characteristics listed in Table 1) were entered as covariates in a logistic regression model to identify which characteristics were associated with thrombophilic factors. Two-tailed p values less than 0.05 were considered significant.
| Results |
|---|
|
|
|---|
In univariate analysis (
2 test) the presence of systolic ventricular dysfunction (left ventricular ejection fraction <0.50), history of acute myocardial infarction, history of chronic obstructive pulmonary disease, and history of deep venous thrombosis or pulmonary embolism associated with the presence of thrombophilic factors (Table 1). Multivariate analysis (logistic regression analysis, see "Statistical Analysis" for details) identified age, history of acute myocardial infarction, and history of chronic obstructive pulmonary disease to associate with the presence of thrombophilic factors (Table 1).
Thrombophilic Variables and Thrombin Generation, Procoagulant Activity of Thrombin, or Fibrinolysis
Patients with at least one, with more than one, or more than two thrombophilic factors had higher preoperative F1+2 levels than patients without thrombophilic factors (median [interquartile range] 0.55 [0.34] vs 0.45 [0.21] nmol/L, p = 0.009; 0.55 [0.25] vs 0.45 [0.21] nmol/L, p = 0.020; and 0.66 [0.26] vs 0.45 (0.21) nmol/L, p = 0.020, respectively). However, F1+2, D-dimer, or SFC levels during CPB and the postoperative period were not different in patients with or without thrombophilic variables (repeated measures ANOVA, p = 0.09, p = 0.08, and p = 0.60, respectively) (Fig 1, panels A to C). Also, the patients with more than one thrombophilic factor (n = 11) did not have higher F1+2, D-dimer, or SFC levels than others (p = 0.09, p = 0.05, and p = 0.69, respectively) (data shown for F1+2, Fig 2, panel A). The same was true for patients with more than two thrombophilic variables (n = 5) (p = 0.67, p = 0.06, and p = 0.11 for F1+2, D-dimer, and SFC, respectively; data not shown). Similar analyses were made separately for each thrombophilic variable tested. In these analyses none of the thrombophilic variables associated with higher perioperative levels of F1+2, D-dimer, or SFC (F1+2 data shown of patients with low levels of the natural anticoagulants antithrombin, protein C, or protein S and of patients with APC resistance due to factor V Leiden [Fig 2, panels B and C]).
|
|
| Comment |
|---|
|
|
|---|
Exposure of blood to the artificial surfaces of the extracorporeal circuit causes a burst in thrombin generation when CPB is initiated [13], but the surgical wound is a major source of tissue factor-mediated thrombin generation during CPB [19]. Thrombin generation in pericardial blood during CABG is fulminant [20] and the pericardial wound contains high concentrations of soluble tissue factor and procoagulant microparticles [21, 22]. Microparticle-bound tissue factor obtained from pericardial blood during CPB stimulates thrombin formation [23], and also microparticle-free soluble tissue factor from pericardial blood is capable of activating factor VII in the presence of stimulated or nonstimulated monocytes and to a lesser extent in the presence of platelets [24]. Avoiding reinfusion of blood aspirated from the surgical field has been shown to attenuate CPB-associated activation of coagulation [25–27]. Processing of aspirated blood with a cell saver also reduces CPB-induced thrombin generation [28] and has been advocated as one method to limit the thrombotic response associated with open heart surgery [19]. Cardiotomy suction without cell saver processing was used in our study and clearly contributed to the observed elevations of markers of thrombin generation and activity. Therefore, we cannot rule out the possibility that any potential modest association between thrombophilic factors and the measured thrombin markers during CPB remained undetected because of the study design.
Abnormal thrombophilia screening was common in our patients, as almost half were identified with at least one thrombophilic factor. However, our definition of thrombophilic factors included less stringent criteria than usual with regard to antithrombin and factor VIII, which were considered abnormal if they were in the lowest decile or above the reference range, respectively; rather than being below 60% or above 200%, respectively. None of our patients had factor VIII above 200% and the three patients with antithrombin below 60% did not differ from others with regard to activation of coagulation during CPB. Nevertheless, our finding of any thrombophilic abnormality in 44% of the patients is consistent with the reported prevalence of thrombophilia in patients with peripheral arterial disease and carotid artery stenosis [1–3]. Interestingly, in our study a history of myocardial infarction or chronic pulmonary disease associated with the presence of thrombophilic factors. Antithrombin activity was below the reference level in a large proportion of our patients with antithrombin levels suggestive of a modest acquired antithrombin deficiency even though our patients had not received heparin treatment preoperatively [29]. Results from previous large cross-sectional studies assessing the association between antithrombin level and prevalent coronary artery disease have been contradictory [30, 31]. However, prospective studies have shown that antithrombin level associates with the risk of cardiac events [32, 33]. Previously, antithrombin activity was measured in our hospital with the same method as in the present study, in patients with acute coronary syndrome [34]. Low antithrombin level in the acute phase associated with the risk of cardiovascular events, but antithrombin increased and was normal in a later stable phase of coronary artery disease [34]. Theoretically, lowered preoperative levels of antithrombin could reflect impaired thrombin control on endothelial surfaces and clinically impair the efficacy of heparin during CPB. However, in the present study there was no evidence of increased thrombin generation or procoagulant activity perioperatively in patients with even the lowest antithrombin levels.
There are some limitations to the present study. While the sample size was sufficient for studying the effect of thrombophilia in general on perioperative activation of coagulation, it was insufficient for evaluation of the impact of individual rare coagulation abnormalities. There was also no angiographic evaluation of graft patency postoperatively in our study. Previously, APC resistance or preoperative measurements of coagulation factor levels have not been shown to associate with graft patency [35–37]. While graft patency is a relevant outcome measure after CABG, perioperative thrombin generation, which was measured in the present study, may also be clinically significant as it has been shown to associate with postoperative myocardial damage [14], postoperative hemodynamic recovery [14, 15], and other markers of postoperative organ dysfunction [15].
In conclusion, we demonstrated that thrombophilic markers do not associate with increased perioperative generation of thrombin or its procoagulant activity in patients undergoing CABG. Our results do not support the use of routine preoperative thrombophilia screening before CABG.
| Acknowledgments |
|---|
|
|
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
P. Raivio, R. Lassila, and J. Petaja Thrombin in myocardial ischemia-reperfusion during cardiac surgery. Ann. Thorac. Surg., July 1, 2009; 88(1): 318 - 325. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. Colman Invited Commentary Ann. Thorac. Surg., February 1, 2008; 85(2): 542 - 542. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |