Domenico Paparella, MD, Crescenzia Rotunno, BS, Micaela De Palo, MD, Simona Finamore, MD, Pietro Guida, PhD, Gianni Rubino, MD, Luigi de Luca Tupputi Schinosa, MD, and Tommaso Fiore, MD Division of Cardiac Surgery and Division of Anesthesia, Department of Emergency and Organ Transplant, University of Bari Aldo Moro, Bari, Italy
Background. Antithrombin (AT) concentrations are reduced after cardiac surgery with cardiopulmonary bypass compared with the preoperative levels. Low postoperative AT is associated with worse short- and mid-term clinical outcomes. The aim of the study is to evaluate the effects of AT administration on activation of the coagulation and fibrinolytic systems, platelet function, and the inflammatory response in patients with low postoperative AT levels. Methods. Sixty patients with postoperative AT levels of less than 65% were randomly assigned to receive purified AT (5000 IU in three administrations) or placebo in the postoperative intensive care unit. Thirty patients with postoperative AT levels greater than 65% were observed as controls. Interleukin 6 (a marker of inflammation), prothrombin fragment 1-2 (a marker of thrombin generation), plasmin–antiplasmin complex (a marker of fibrinolysis), and platelet factor 4 (a marker of platelet activation) were measured at six different times.
Results. Compared with the no AT group and control patients, patients receiving AT showed significantly higher AT values until 48 hours after the last administration. Analysis of variance for repeated measures showed a significant effect of study treatment in reducing prothrombin fragment 1-2 (p [ 0.009; interaction with time sample, p [ 0.006) and plasmin–antiplasmin complex (p < 0.001; interaction with time sample, p < 0.001) values but not interleukin 6 (p [ 0.877; interaction with time sample, p [ 0.521) and platelet factor 4 (p [ 0.913; interaction with time sample, p [ 0.543). No difference in chest tube drainage, reopening for bleeding, and blood transfusion was observed. Conclusions. Antithrombin administration in patients with low AT activity after surgery with cardiopulmonary bypass reduces postoperative thrombin generation and fibrinolysis with no effects on platelet activation and inflammatory response.
P
low AT levels after cardiac surgery are associated with worse short- and mid-term clinical outcomes [5–7]. It remains unclear, however, whether low AT levels after surgery contribute to unfavorable outcomes or should be considered another biomarker indicating worse general clinical conditions. Therefore this study evaluated the effects of AT administration on activation of the coagulation and fibrinolytic systems, platelet function, and the inflammatory response of patients undergoing cardiac surgery with CPB who have low postoperative AT levels. We hypothesize that normalization of postoperative AT concentration may reduce hemostatic and inflammatory alterations with consequent possible advantages on postoperative blood loss, blood transfusion, and organ damage after surgery.
atients undergoing cardiac surgery experience important alterations of the coagulation and inflammatory systems. Surgical trauma, the use of cardiopulmonary bypass (CPB), and ischemia-reperfusion are the most relevant factors leading to a peculiar hematologic state that may be associated with complicated postoperative outcome [1]. Cardiopulmonary bypass requires massive anticoagulation, but despite heparin administration and elevated activated clotting time, the coagulation system is activated and thrombin generation is not abolished [2, 3]. Antithrombin (AT) levels are reduced after cardiac surgery compared with the preoperative state [4], and the main factors associated with low postoperative AT levels are age, CPB duration, and preoperative AT levels. Different studies have demonstrated that
(Ann Thorac Surg 2014;97:1207–13) Ó 2014 by The Society of Thoracic Surgeons
Accepted for publication Nov 18, 2013. Address correspondence to Dr Paparella, Department of Emergency and Organ Transplant, Division of Cardiac Surgery, University of Bari Aldo Moro, Piazza Giulio Cesare 11, 70100 Bari, Italy; e-mail: domenico. [email protected].
Ó 2014 by The Society of Thoracic Surgeons Published by Elsevier Inc
Material and Methods The Ethic Committee of the Policlinico of Bari University Hospital approved the study, and all patients signed their 0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2013.11.040
ADULT CARDIAC
Antithrombin Administration in Patients With Low Antithrombin Values After Cardiac Surgery: A Randomized Controlled Trial
1208
PAPARELLA ET AL ANTITHROMBIN AFTER CARDIAC SURGERY
informed consent. The protocol was properly registered at a public trials registry, www.clinicaltrial.gov (trial identifier NCT01201070). ADULT CARDIAC
Patients Patients were enrolled from September 2009 to October 2012. Patients who were candidates for cardiac operations with CPB were eligible to be enrolled in the study except for the following exclusion criteria: (1) a positive history for allergic reactions to AT; (2) off-pump cardiac surgery; (3) AT administration during surgery or within 48 hours before the operation (Antithrombin administration before CPB occurred in the event of heparin resistance [activated clotting time < 400 seconds after administration of heparin 300 U/kg and subsequent adjunctive heparin bolus]. Patients with heparin resistance were excluded from the study.); (4) treatment with steroids or nonsteroidal antiinflammatory drugs within 48 hours before the operation; (5) preoperative known coagulation disorders; (6) platelet count of less than 30 103/mL; (7) chronic renal failure or preoperative dialysis treatment; (8) severe liver failure; (9) participation in another trial in the last 6 months; (10) deep hypothermic circulatory arrest during surgery; (11) emergency operation; (12) reintervention; (13) CPB time longer than 180 minutes; and (14) subjects unable to give consent. The design of the study is prospective, randomized, controlled, and double blind. The randomization was computer generated. Consenting patients were screened before surgery, and randomization took place once the postoperative AT levels were measured immediately after the end of surgery: patients with postoperative AT levels of less than 65% were randomly assigned to either the AT group (treatment with AT) or the no AT group (no treatment with AT). The intensive care unit (ICU) anesthetic, surgical, and nursing team was not aware of the type of treatment. Patients (n ¼ 77) with postoperative levels of AT greater than 65% were allocated into a control group. To enroll 60 patients with an AT of less than 65%, a total of 137 consenting patients had to be scrutinized after surgery. Only the first 30 control group patients were selected for data analysis to make the three groups uniform in number. TREATMENT. Patients were randomly assigned to receive AT (CSL Behring, Kedrion, Grifols, Baxter) or an equal volume of saline solution. Antithrombin was administered in the AT group as follows: 3,000 IU given as a bolus at the time of randomization, 1,000 IU after 8 hours, and 1,000 IU after 16 hours (total, 5,000/UI in 24 hours). Saline solution was administered intravenously in the no AT group patients in the same time. DATA COLLECTION. Preoperative patient characteristics, basic laboratory results, and clinical outcomes were recorded. DEFINITIONS. Obesity was defined as body mass index exceeding 30 kg/m2. Hypertension was defined as blood pressure exceeding 140/90 mm Hg, having a history of high blood pressure, or needing antihypertensive medications. Patients with a history of diabetes regardless of
Ann Thorac Surg 2014;97:1207–13
the duration of disease (or the need for antidiabetic agents) were considered diabetic. Chronic obstructive pulmonary disease was defined as forced expiratory volume of air in 1 second less than 75% or the need for pharmacologic therapy for the treatment of chronic pulmonary compromise. Preoperative renal disease was defined as serum creatinine greater than 2.0 mg/dL. Death within the same hospital admission regardless of the cause was defined as operative mortality. Cerebrovascular disease was regarded as any transient ischemia attack, reversible ischemic neurologic deficit, or stroke. Acute renal failure was defined as the following: new onset of postoperative creatinine greater than 2.0 mg/dL; increase of creatinine greater than twofold compared with preoperative creatinine; or requirement for postoperative dialysis. During and after the operation, blood and blood products were transfused only if the patient had signs of hypovolemia (hypotension or tachycardia) and according with the following criteria: allogeneic packed red blood cells were transfused if the hemoglobin value was less than 8 g/dL or the hematocrit was less than 24%. Fresh-frozen plasma was infused if the prothrombin time–international normalized ratio (PT-INR) value after protamine administration was at least 1.5 times the baseline in the presence of significant bleeding, and platelet concentrates were transfused with a platelet count of less than 50 103/mL.
Laboratory Variables ACTIVATION OF THE COAGULATION SYSTEM. Antithrombin levels, prothrombin time, activated partial thromboplastin time, and fibrinogen levels were measured. Activation of the coagulation system was assessed by measuring thrombin generation, in particular by plasma levels of prothrombin fragment 1-2 (PF-1.2), using commercially available enzyme-linked immunosorbent assay (ELISA) kits. PLATELET ACTIVATION. Platelet function and activation was measured indirectly as a drop in platelet counts and as a direct release of platelet factor 4 (PF-4) in plasma. FIBRINOLYSIS ACTIVATION. Activation of fibrinolysis was assessed by measuring plasma levels of plasmin–antiplasmin (PAP) complex. INFLAMMATORY RESPONSE. Activation of inflammation was assessed by measuring the levels of interleukin 6 (IL-6) and C-reactive protein. MEASUREMENTS OF ORGAN DAMAGE. Levels of cardiac troponin I, as an index of myocardial injury, and creatinine levels, as an index of renal damage, were recorded. SAMPLE COLLECTION. Blood sample collection was carried out from a central venous catheter or peripheral vein. The blood obtained was introduced into blood specimen collection tubes containing sodium citrate 3.8% (for the evaluation of markers of coagulation and fibrinolysis) or citrate theophylline adenosine and dipyridamole (for evaluation of markers of platelet activation), kept in ice for 15 minutes, and then centrifuged at 3,500 rpm for 15 minutes at 4 C. The plasma thus obtained was stored at 80 C. In each period of observation, the hematocrit
Ann Thorac Surg 2014;97:1207–13
Perioperative Management After premedication with lorazepam, anesthesia was induced with a combination of fentanyl, midazolam, and sodium thiopental and maintained with propofol. Heparin was injected as a single bolus, and anticoagulation was monitored by the activated clotting time (Hemochron 8; ITC, Pleasanton, CA). Additional heparin (5,000 U) was given if the activated clotting time fell to less than 400 seconds. After surgery, protamine sulfate was administered to neutralize heparin (1 mg of protamine/100 IU of heparin). Tranexamic acid (3 g) was administered in all patients. Cell-saving devices were not used. In the ICU, patients were extubated when they were hemodynamically stable, fully rewarmed, awake, without surgical bleeding, and with optimal blood gases. Chest tubes were removed when drainage was less than 10 mL/h for at least 4 hours, which generally occurred within 24 to 48 hours. Standard treatment with regard to anticoagulation, antibiotic therapy, and intensive care protocols were applied to all patients.
Statistical Analysis The data are given as mean values, with standard deviation or standard error, and categorical variables are reported as a percentage. The principal aim of this trial was to evaluate the effects of the administration of AT on activation of the coagulation and inflammatory systems (PF-1.2, IL-6, PAP, and PF4). Statistical analyses were performed using two-factor repeated-measures analysis of variance with one repeated factor (time sampling) and one grouping factor (AT versus no AT groups). Considering five postoperative repeated measures (coefficient of correlation, 0.30), it was calculated that to get an 80% chance of detecting a standardized mean difference of 50%, at least 28 patients per group were needed. Variables were compared using the median test or Wilcoxon rank-sum test for two independent samples. Frequencies were compared by Fisher’s exact test. Comparisons of the two study groups with control patients were made for
1209
exploratory purposes. The statistical analyses and sample size calculation were performed using Stata 12 (StataCorp, College Station, TX), and probability values less than 0.05 were considered statistically significant.
Results Patients were screened at the ICU arrival after cardiac surgery. The first 30 patients with AT levels greater than 65% were retained as a control group. The remaining 60 patients had AT levels less than 65% and were randomly assigned to receive AT administration or placebo. Baseline characteristics and intraoperative variables are shown in Table 1. Groups were comparable for preoperative and intraoperative data with the exception of age, which was lower in control patients than those in the no AT group (p ¼ 0.040), and history of myocardial infarction, which was more frequent in the no AT group than in the AT group (p ¼ 0.026). Among preoperative laboratory data, fibrinogen was higher in control patients than in the AT group (p ¼ 0.019). Operative details were not different among groups (Table 1). Figure 1 shows AT levels as a function of different sample times. Preoperative AT was significantly higher in the control group than the other groups, without differences between the AT and no AT groups. At ICU arrival before randomization time, control patients had greater AT values than those in the other groups (randomization exclusion criteria) without differences between the AT and no AT groups. After randomization until 48 hours after the last AT administration, there were differences among all groups, with patients in the AT group showing significantly higher values than the control and No AT groups (p < 0.001 at each time sample). The no AT group had lower values than the control group at 90 minutes after the first administration of AT (p ¼ 0.002). On the fifth postoperative day, the three groups had similar AT values (p ¼ 0.210). Figure 2 displays PF-1.2, PAP, IL-6, and PF4 mean values with standard error according to study groups at specific times. At ICU arrival greater values were observed than those measured preoperatively without a significant difference among the groups. At 90 minutes after the first administration of AT, PF-1.2 significantly decreased in each study group (p < 0.001), and PAP significantly decreased only in the AT group (p < 0.001), whereas no difference was detected for IL-6 and PF4. Analysis of variance for repeated measures applied to the AT versus no AT group for postrandomization samples from 90 minutes after the first administration of AT to the fifth postoperative day showed a statistically significant time sample change for all variables (p < 0.001 for each variable). Study treatment showed a significant effect, reducing PF-1.2 (p ¼ 0.009; interaction with time sample, p ¼ 0.006) and PAP (p < 0.001; interaction with time sample, p < 0.001) values but not IL-6 (p ¼ 0.877; interaction with time sample, p ¼ 0.521) and PF4 (p ¼ 0.913; interaction with time sample, p ¼ 0.543). Between groups differences were detected for PF-1.2 (p ¼ 0.006) and PAP at 90 minutes after the first AT administration
ADULT CARDIAC
was measured and results obtained were corrected for hemodilution multiplying the assay by the ratio between baseline hematocrit and the hematocrit measured in the sample. Blood was sampled at the following times: (1) preoperatively in the morning of surgery, before induction of anesthesia; (2) baseline, ICU arrival; (3) 90 minutes after the last AT administration; (4) 1 hour after the last AT administration; (5) 24 hours after the last AT administration; (6) 48 hours after the last AT administration; and (7) 5 days after surgery. ASSAYS. Plasma was assayed by monoclonal antibody sandwich ELISA technique: IL-6 (Human Interleukin-6 EH2IL6, Thermo Scientific, Vernon Hills, IL); PF-1.2 (Enzygnost F1þ2 Monoclonal, Siemens Health Care, Munich, Germany); PAP (PAP micro ELISA Kit, DRG Instruments GmbH, Germany); and PF4 (Imuclone Platelet Factor 4 ELISA Kit, American Diagnostica Inc, Stamford, CT). All ELISA assays were double tested, and the mean value was used for analysis.
PAPARELLA ET AL ANTITHROMBIN AFTER CARDIAC SURGERY
1210
PAPARELLA ET AL ANTITHROMBIN AFTER CARDIAC SURGERY
Ann Thorac Surg 2014;97:1207–13
Table 1. Preoperative Patients Characteristics According to the Study Groupsa
ADULT CARDIAC
Variable
All (n ¼ 90)
Controls (n ¼ 30)
No AT (n ¼ 30)
AT (n ¼ 30)
p Value
Male sex Age (y) Body mass index (kg/m2) Obesity Arterial hypertension Diabetes Dyslipidemia History of myocardial infarction Unstable angina Left ventricular ejection fraction Atrial fibrillation Neurologic dysfunction Renal dysfunction Dialysis Hematocrit (%) aPTT (s) PT (s) Fibrinogen (mg/dL) Platelets ( 103/mL) CRP (mg/L) Troponin I (ng/mL) Creatinine (mg/dL) GOT (U/L) GPT (U/L) Operative details CABG Aortic valve replacement Mitral valve surgery Isolated thoracic aortic surgery Root replacement Intervention duration (hours) CPB duration (min) Cross-clamp time (min)
72% 67 10 28 5 29% 76% 26% 39% 13% 9% 0.50 0.09 22% 4% 9% 0% 39.2 4.9 33 6 14.1 1.7 360 96 216 62 9 13 0.09 0.43 1.03 0.31 22 11 38 18
73% 65 12 28 6 27% 67% 23% 53% 10% 10% 0.53 0.09 30% 7% 7% 0% 38.9 4.1 35 5 14.6 1.4 395 123 215 62 10 12 0.06 0.19 0.95 0.27 23 15 45 27
73% 70 10 28 5 27% 77% 33% 37% 27% 13% 0.49 0.11 13% 0% 13% 0% 39.4 5.7 32 6 14 1.7 356 85 229 61 8 13 0.18 0.71 1.1 0.33 21 8 35 9
70% 67 9 29 4 33% 83% 20% 27% 3% 3% 049 0.08 23% 7% 7% 0% 39.1 4.7 30 5 13.8 2 329 60 205 62 8 16 0.04 0.07 1.04 0.33 20 9 33 11
1.000 0.027 0.282 0.876 0.360 0.565 0.126 0.035 0.522 0.198 0.336 0.540 0.722 – 0.834 0.056 0.301 0.008 0.128 0.276 0.730 0.422 0.669 0.288
36% 28% 32% 6% 13% 4. 4 1.0 123 39 91 36
37% 37% 33% 7% 7% 4.5 1.1 120 39 88 39
43% 20% 27% 10% 17% 4.5 0.9 123 39 91 37
27% 27% 37% 0% 17% 4.3 0.9 128 39 95 33
0.438 0.390 0.782 0.363 0.470 0.654 0.988 0.904
a
Values are reported as mean standard deviation or percentage.
aPTT ¼ activated partial thromboplastin time; AT ¼ antithrombin; CABG ¼ coronary artery bypass grafting; CPB ¼ cardiopulmonary bypass; CRP ¼ C-reactive protein; GOT ¼ glutamate oxaloacetate transaminase; GPT ¼ glutamate pyruvate transaminase; PT ¼ prothrombin time.
(p < 0.001), and at 48 hours after the last AT administration for PF-1.2 (p < 0.001) with lower values in treated patients in comparison to nontreated. Control patients had lower values of PF-1.2 at 90 minutes after the first AT administration (p < 0.001) than the no AT patients and greater values at the fifth postoperative day than those in the AT group (p ¼ 0.020). Moreover, AT patients had PAP values lower than control patients at 90 minutes after the first AT administration and at 48 hours after the last AT administration (respectively, p ¼ 0.001 and p ¼ 0.014). Control patients had greater IL-6 values 90 minutes after the first AT administration in comparison to the other groups (p < 0.001 versus AT and p ¼ 0.035 versus no AT). Table 2 shows intraoperative and postoperative variables according to study groups. Before randomization, fibrinogen was significantly higher in the control group
than the other groups. After randomization, fibrinogen and glutamate pyruvate transaminase were greater in the control group than the other groups at 1 hour after the first AT administration. Postoperative complications were not different among groups. No difference in chest tube drainage, reopening for bleeding, and blood transfusion was observed. Only the onset of atrial fibrillation was lower in the control group than in the no AT group. Cardiac troponin I and creatinine were similar among groups.
Comment Antithrombin is a serine protease inhibitor synthesized by the liver. Together with protein C and protein S, it exerts a role as an endogenous anticoagulant. Because
Fig 1. Antithrombin (AT) levels in control patients (Controls; black bars), patients receiving AT (white bars), and those receiving placebo (no AT; gray bars), measured preoperatively (Pre-op), on intensive care unit acceptance (ICU), 90 minutes after the first AT administration (900 AT), 1 hour after the last AT administration (1h AT), 24 hours after the last AT administration (24h AT), 48 hours after the last AT administration (48h AT), and on the fifth postoperative day (day 5). Values are reported as mean standard error.
of catalysis by heparin, its activity is enhanced. Patients undergoing cardiac surgery experience a significant decrease of AT values compared with preoperative levels [4–7], and lower postoperative AT levels have a negative prognostic influence, being independent predictors of several postoperative complications: prolonged ventilation time, low cardiac output syndrome, atrial fibrillation, increased incidence of reintervention for bleeding, need of blood product transfusions, neurological
PAPARELLA ET AL ANTITHROMBIN AFTER CARDIAC SURGERY
1211
complications, and ICU and overall length of stay [5–7]. Currently, AT administration during cardiac operations, especially before CPB establishment in case of heparin resistance, has become a well-established practice [8]. Recently, Ranucci and colleagues [9], in a prospective randomized trial, evaluated the effects of preoperative AT administration to reach an AT activity value of 120% in patients undergoing cardiac surgery with CPB. They showed that preoperative AT administration prevents heparin resistance before CPB establishment and favors higher postoperative AT values, but no clinical benefit could be demonstrated. Postoperative AT administration with the aim of restoring physiologic values is still a therapeutic strategy whose biologic effects and potential clinical benefit have not been documented. Guidelines place AT supplementation for the prevention of thromboembolic complications in selected patient populations in class IIB [10]. For this reason we investigated for the first time the effects of postoperative AT administration in patients with low AT levels after cardiac surgery. Moreover, we analyzed patients with higher AT levels after surgery who served as control subjects to be compared with those undergoing randomization. As previously demonstrated [6], low postoperative AT levels were significantly influenced by preoperative AT levels; in fact, randomized patients had significantly lower AT preoperative values compared with control patients. The cutoff value used to select patients to undergo randomization (
Background. Antithrombin (AT) concentrations are reduced after cardiac surgery with cardiopulmonary bypass compared with the preoperative levels. Low postoperative AT is associated with worse short- and mid-term clinical outcomes. The aim of the study is to evaluate the effects of AT administration on activation of the coagulation and fibrinolytic systems, platelet function, and the inflammatory response in patients with low postoperative AT levels. Methods. Sixty patients with postoperative AT levels of less than 65% were randomly assigned to receive purified AT (5000 IU in three administrations) or placebo in the postoperative intensive care unit. Thirty patients with postoperative AT levels greater than 65% were observed as controls. Interleukin 6 (a marker of inflammation), prothrombin fragment 1-2 (a marker of thrombin generation), plasmin–antiplasmin complex (a marker of fibrinolysis), and platelet factor 4 (a marker of platelet activation) were measured at six different times.
Results. Compared with the no AT group and control patients, patients receiving AT showed significantly higher AT values until 48 hours after the last administration. Analysis of variance for repeated measures showed a significant effect of study treatment in reducing prothrombin fragment 1-2 (p [ 0.009; interaction with time sample, p [ 0.006) and plasmin–antiplasmin complex (p < 0.001; interaction with time sample, p < 0.001) values but not interleukin 6 (p [ 0.877; interaction with time sample, p [ 0.521) and platelet factor 4 (p [ 0.913; interaction with time sample, p [ 0.543). No difference in chest tube drainage, reopening for bleeding, and blood transfusion was observed. Conclusions. Antithrombin administration in patients with low AT activity after surgery with cardiopulmonary bypass reduces postoperative thrombin generation and fibrinolysis with no effects on platelet activation and inflammatory response.
P
low AT levels after cardiac surgery are associated with worse short- and mid-term clinical outcomes [5–7]. It remains unclear, however, whether low AT levels after surgery contribute to unfavorable outcomes or should be considered another biomarker indicating worse general clinical conditions. Therefore this study evaluated the effects of AT administration on activation of the coagulation and fibrinolytic systems, platelet function, and the inflammatory response of patients undergoing cardiac surgery with CPB who have low postoperative AT levels. We hypothesize that normalization of postoperative AT concentration may reduce hemostatic and inflammatory alterations with consequent possible advantages on postoperative blood loss, blood transfusion, and organ damage after surgery.
atients undergoing cardiac surgery experience important alterations of the coagulation and inflammatory systems. Surgical trauma, the use of cardiopulmonary bypass (CPB), and ischemia-reperfusion are the most relevant factors leading to a peculiar hematologic state that may be associated with complicated postoperative outcome [1]. Cardiopulmonary bypass requires massive anticoagulation, but despite heparin administration and elevated activated clotting time, the coagulation system is activated and thrombin generation is not abolished [2, 3]. Antithrombin (AT) levels are reduced after cardiac surgery compared with the preoperative state [4], and the main factors associated with low postoperative AT levels are age, CPB duration, and preoperative AT levels. Different studies have demonstrated that
(Ann Thorac Surg 2014;97:1207–13) Ó 2014 by The Society of Thoracic Surgeons
Accepted for publication Nov 18, 2013. Address correspondence to Dr Paparella, Department of Emergency and Organ Transplant, Division of Cardiac Surgery, University of Bari Aldo Moro, Piazza Giulio Cesare 11, 70100 Bari, Italy; e-mail: domenico. [email protected].
Ó 2014 by The Society of Thoracic Surgeons Published by Elsevier Inc
Material and Methods The Ethic Committee of the Policlinico of Bari University Hospital approved the study, and all patients signed their 0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2013.11.040
ADULT CARDIAC
Antithrombin Administration in Patients With Low Antithrombin Values After Cardiac Surgery: A Randomized Controlled Trial
1208
PAPARELLA ET AL ANTITHROMBIN AFTER CARDIAC SURGERY
informed consent. The protocol was properly registered at a public trials registry, www.clinicaltrial.gov (trial identifier NCT01201070). ADULT CARDIAC
Patients Patients were enrolled from September 2009 to October 2012. Patients who were candidates for cardiac operations with CPB were eligible to be enrolled in the study except for the following exclusion criteria: (1) a positive history for allergic reactions to AT; (2) off-pump cardiac surgery; (3) AT administration during surgery or within 48 hours before the operation (Antithrombin administration before CPB occurred in the event of heparin resistance [activated clotting time < 400 seconds after administration of heparin 300 U/kg and subsequent adjunctive heparin bolus]. Patients with heparin resistance were excluded from the study.); (4) treatment with steroids or nonsteroidal antiinflammatory drugs within 48 hours before the operation; (5) preoperative known coagulation disorders; (6) platelet count of less than 30 103/mL; (7) chronic renal failure or preoperative dialysis treatment; (8) severe liver failure; (9) participation in another trial in the last 6 months; (10) deep hypothermic circulatory arrest during surgery; (11) emergency operation; (12) reintervention; (13) CPB time longer than 180 minutes; and (14) subjects unable to give consent. The design of the study is prospective, randomized, controlled, and double blind. The randomization was computer generated. Consenting patients were screened before surgery, and randomization took place once the postoperative AT levels were measured immediately after the end of surgery: patients with postoperative AT levels of less than 65% were randomly assigned to either the AT group (treatment with AT) or the no AT group (no treatment with AT). The intensive care unit (ICU) anesthetic, surgical, and nursing team was not aware of the type of treatment. Patients (n ¼ 77) with postoperative levels of AT greater than 65% were allocated into a control group. To enroll 60 patients with an AT of less than 65%, a total of 137 consenting patients had to be scrutinized after surgery. Only the first 30 control group patients were selected for data analysis to make the three groups uniform in number. TREATMENT. Patients were randomly assigned to receive AT (CSL Behring, Kedrion, Grifols, Baxter) or an equal volume of saline solution. Antithrombin was administered in the AT group as follows: 3,000 IU given as a bolus at the time of randomization, 1,000 IU after 8 hours, and 1,000 IU after 16 hours (total, 5,000/UI in 24 hours). Saline solution was administered intravenously in the no AT group patients in the same time. DATA COLLECTION. Preoperative patient characteristics, basic laboratory results, and clinical outcomes were recorded. DEFINITIONS. Obesity was defined as body mass index exceeding 30 kg/m2. Hypertension was defined as blood pressure exceeding 140/90 mm Hg, having a history of high blood pressure, or needing antihypertensive medications. Patients with a history of diabetes regardless of
Ann Thorac Surg 2014;97:1207–13
the duration of disease (or the need for antidiabetic agents) were considered diabetic. Chronic obstructive pulmonary disease was defined as forced expiratory volume of air in 1 second less than 75% or the need for pharmacologic therapy for the treatment of chronic pulmonary compromise. Preoperative renal disease was defined as serum creatinine greater than 2.0 mg/dL. Death within the same hospital admission regardless of the cause was defined as operative mortality. Cerebrovascular disease was regarded as any transient ischemia attack, reversible ischemic neurologic deficit, or stroke. Acute renal failure was defined as the following: new onset of postoperative creatinine greater than 2.0 mg/dL; increase of creatinine greater than twofold compared with preoperative creatinine; or requirement for postoperative dialysis. During and after the operation, blood and blood products were transfused only if the patient had signs of hypovolemia (hypotension or tachycardia) and according with the following criteria: allogeneic packed red blood cells were transfused if the hemoglobin value was less than 8 g/dL or the hematocrit was less than 24%. Fresh-frozen plasma was infused if the prothrombin time–international normalized ratio (PT-INR) value after protamine administration was at least 1.5 times the baseline in the presence of significant bleeding, and platelet concentrates were transfused with a platelet count of less than 50 103/mL.
Laboratory Variables ACTIVATION OF THE COAGULATION SYSTEM. Antithrombin levels, prothrombin time, activated partial thromboplastin time, and fibrinogen levels were measured. Activation of the coagulation system was assessed by measuring thrombin generation, in particular by plasma levels of prothrombin fragment 1-2 (PF-1.2), using commercially available enzyme-linked immunosorbent assay (ELISA) kits. PLATELET ACTIVATION. Platelet function and activation was measured indirectly as a drop in platelet counts and as a direct release of platelet factor 4 (PF-4) in plasma. FIBRINOLYSIS ACTIVATION. Activation of fibrinolysis was assessed by measuring plasma levels of plasmin–antiplasmin (PAP) complex. INFLAMMATORY RESPONSE. Activation of inflammation was assessed by measuring the levels of interleukin 6 (IL-6) and C-reactive protein. MEASUREMENTS OF ORGAN DAMAGE. Levels of cardiac troponin I, as an index of myocardial injury, and creatinine levels, as an index of renal damage, were recorded. SAMPLE COLLECTION. Blood sample collection was carried out from a central venous catheter or peripheral vein. The blood obtained was introduced into blood specimen collection tubes containing sodium citrate 3.8% (for the evaluation of markers of coagulation and fibrinolysis) or citrate theophylline adenosine and dipyridamole (for evaluation of markers of platelet activation), kept in ice for 15 minutes, and then centrifuged at 3,500 rpm for 15 minutes at 4 C. The plasma thus obtained was stored at 80 C. In each period of observation, the hematocrit
Ann Thorac Surg 2014;97:1207–13
Perioperative Management After premedication with lorazepam, anesthesia was induced with a combination of fentanyl, midazolam, and sodium thiopental and maintained with propofol. Heparin was injected as a single bolus, and anticoagulation was monitored by the activated clotting time (Hemochron 8; ITC, Pleasanton, CA). Additional heparin (5,000 U) was given if the activated clotting time fell to less than 400 seconds. After surgery, protamine sulfate was administered to neutralize heparin (1 mg of protamine/100 IU of heparin). Tranexamic acid (3 g) was administered in all patients. Cell-saving devices were not used. In the ICU, patients were extubated when they were hemodynamically stable, fully rewarmed, awake, without surgical bleeding, and with optimal blood gases. Chest tubes were removed when drainage was less than 10 mL/h for at least 4 hours, which generally occurred within 24 to 48 hours. Standard treatment with regard to anticoagulation, antibiotic therapy, and intensive care protocols were applied to all patients.
Statistical Analysis The data are given as mean values, with standard deviation or standard error, and categorical variables are reported as a percentage. The principal aim of this trial was to evaluate the effects of the administration of AT on activation of the coagulation and inflammatory systems (PF-1.2, IL-6, PAP, and PF4). Statistical analyses were performed using two-factor repeated-measures analysis of variance with one repeated factor (time sampling) and one grouping factor (AT versus no AT groups). Considering five postoperative repeated measures (coefficient of correlation, 0.30), it was calculated that to get an 80% chance of detecting a standardized mean difference of 50%, at least 28 patients per group were needed. Variables were compared using the median test or Wilcoxon rank-sum test for two independent samples. Frequencies were compared by Fisher’s exact test. Comparisons of the two study groups with control patients were made for
1209
exploratory purposes. The statistical analyses and sample size calculation were performed using Stata 12 (StataCorp, College Station, TX), and probability values less than 0.05 were considered statistically significant.
Results Patients were screened at the ICU arrival after cardiac surgery. The first 30 patients with AT levels greater than 65% were retained as a control group. The remaining 60 patients had AT levels less than 65% and were randomly assigned to receive AT administration or placebo. Baseline characteristics and intraoperative variables are shown in Table 1. Groups were comparable for preoperative and intraoperative data with the exception of age, which was lower in control patients than those in the no AT group (p ¼ 0.040), and history of myocardial infarction, which was more frequent in the no AT group than in the AT group (p ¼ 0.026). Among preoperative laboratory data, fibrinogen was higher in control patients than in the AT group (p ¼ 0.019). Operative details were not different among groups (Table 1). Figure 1 shows AT levels as a function of different sample times. Preoperative AT was significantly higher in the control group than the other groups, without differences between the AT and no AT groups. At ICU arrival before randomization time, control patients had greater AT values than those in the other groups (randomization exclusion criteria) without differences between the AT and no AT groups. After randomization until 48 hours after the last AT administration, there were differences among all groups, with patients in the AT group showing significantly higher values than the control and No AT groups (p < 0.001 at each time sample). The no AT group had lower values than the control group at 90 minutes after the first administration of AT (p ¼ 0.002). On the fifth postoperative day, the three groups had similar AT values (p ¼ 0.210). Figure 2 displays PF-1.2, PAP, IL-6, and PF4 mean values with standard error according to study groups at specific times. At ICU arrival greater values were observed than those measured preoperatively without a significant difference among the groups. At 90 minutes after the first administration of AT, PF-1.2 significantly decreased in each study group (p < 0.001), and PAP significantly decreased only in the AT group (p < 0.001), whereas no difference was detected for IL-6 and PF4. Analysis of variance for repeated measures applied to the AT versus no AT group for postrandomization samples from 90 minutes after the first administration of AT to the fifth postoperative day showed a statistically significant time sample change for all variables (p < 0.001 for each variable). Study treatment showed a significant effect, reducing PF-1.2 (p ¼ 0.009; interaction with time sample, p ¼ 0.006) and PAP (p < 0.001; interaction with time sample, p < 0.001) values but not IL-6 (p ¼ 0.877; interaction with time sample, p ¼ 0.521) and PF4 (p ¼ 0.913; interaction with time sample, p ¼ 0.543). Between groups differences were detected for PF-1.2 (p ¼ 0.006) and PAP at 90 minutes after the first AT administration
ADULT CARDIAC
was measured and results obtained were corrected for hemodilution multiplying the assay by the ratio between baseline hematocrit and the hematocrit measured in the sample. Blood was sampled at the following times: (1) preoperatively in the morning of surgery, before induction of anesthesia; (2) baseline, ICU arrival; (3) 90 minutes after the last AT administration; (4) 1 hour after the last AT administration; (5) 24 hours after the last AT administration; (6) 48 hours after the last AT administration; and (7) 5 days after surgery. ASSAYS. Plasma was assayed by monoclonal antibody sandwich ELISA technique: IL-6 (Human Interleukin-6 EH2IL6, Thermo Scientific, Vernon Hills, IL); PF-1.2 (Enzygnost F1þ2 Monoclonal, Siemens Health Care, Munich, Germany); PAP (PAP micro ELISA Kit, DRG Instruments GmbH, Germany); and PF4 (Imuclone Platelet Factor 4 ELISA Kit, American Diagnostica Inc, Stamford, CT). All ELISA assays were double tested, and the mean value was used for analysis.
PAPARELLA ET AL ANTITHROMBIN AFTER CARDIAC SURGERY
1210
PAPARELLA ET AL ANTITHROMBIN AFTER CARDIAC SURGERY
Ann Thorac Surg 2014;97:1207–13
Table 1. Preoperative Patients Characteristics According to the Study Groupsa
ADULT CARDIAC
Variable
All (n ¼ 90)
Controls (n ¼ 30)
No AT (n ¼ 30)
AT (n ¼ 30)
p Value
Male sex Age (y) Body mass index (kg/m2) Obesity Arterial hypertension Diabetes Dyslipidemia History of myocardial infarction Unstable angina Left ventricular ejection fraction Atrial fibrillation Neurologic dysfunction Renal dysfunction Dialysis Hematocrit (%) aPTT (s) PT (s) Fibrinogen (mg/dL) Platelets ( 103/mL) CRP (mg/L) Troponin I (ng/mL) Creatinine (mg/dL) GOT (U/L) GPT (U/L) Operative details CABG Aortic valve replacement Mitral valve surgery Isolated thoracic aortic surgery Root replacement Intervention duration (hours) CPB duration (min) Cross-clamp time (min)
72% 67 10 28 5 29% 76% 26% 39% 13% 9% 0.50 0.09 22% 4% 9% 0% 39.2 4.9 33 6 14.1 1.7 360 96 216 62 9 13 0.09 0.43 1.03 0.31 22 11 38 18
73% 65 12 28 6 27% 67% 23% 53% 10% 10% 0.53 0.09 30% 7% 7% 0% 38.9 4.1 35 5 14.6 1.4 395 123 215 62 10 12 0.06 0.19 0.95 0.27 23 15 45 27
73% 70 10 28 5 27% 77% 33% 37% 27% 13% 0.49 0.11 13% 0% 13% 0% 39.4 5.7 32 6 14 1.7 356 85 229 61 8 13 0.18 0.71 1.1 0.33 21 8 35 9
70% 67 9 29 4 33% 83% 20% 27% 3% 3% 049 0.08 23% 7% 7% 0% 39.1 4.7 30 5 13.8 2 329 60 205 62 8 16 0.04 0.07 1.04 0.33 20 9 33 11
1.000 0.027 0.282 0.876 0.360 0.565 0.126 0.035 0.522 0.198 0.336 0.540 0.722 – 0.834 0.056 0.301 0.008 0.128 0.276 0.730 0.422 0.669 0.288
36% 28% 32% 6% 13% 4. 4 1.0 123 39 91 36
37% 37% 33% 7% 7% 4.5 1.1 120 39 88 39
43% 20% 27% 10% 17% 4.5 0.9 123 39 91 37
27% 27% 37% 0% 17% 4.3 0.9 128 39 95 33
0.438 0.390 0.782 0.363 0.470 0.654 0.988 0.904
a
Values are reported as mean standard deviation or percentage.
aPTT ¼ activated partial thromboplastin time; AT ¼ antithrombin; CABG ¼ coronary artery bypass grafting; CPB ¼ cardiopulmonary bypass; CRP ¼ C-reactive protein; GOT ¼ glutamate oxaloacetate transaminase; GPT ¼ glutamate pyruvate transaminase; PT ¼ prothrombin time.
(p < 0.001), and at 48 hours after the last AT administration for PF-1.2 (p < 0.001) with lower values in treated patients in comparison to nontreated. Control patients had lower values of PF-1.2 at 90 minutes after the first AT administration (p < 0.001) than the no AT patients and greater values at the fifth postoperative day than those in the AT group (p ¼ 0.020). Moreover, AT patients had PAP values lower than control patients at 90 minutes after the first AT administration and at 48 hours after the last AT administration (respectively, p ¼ 0.001 and p ¼ 0.014). Control patients had greater IL-6 values 90 minutes after the first AT administration in comparison to the other groups (p < 0.001 versus AT and p ¼ 0.035 versus no AT). Table 2 shows intraoperative and postoperative variables according to study groups. Before randomization, fibrinogen was significantly higher in the control group
than the other groups. After randomization, fibrinogen and glutamate pyruvate transaminase were greater in the control group than the other groups at 1 hour after the first AT administration. Postoperative complications were not different among groups. No difference in chest tube drainage, reopening for bleeding, and blood transfusion was observed. Only the onset of atrial fibrillation was lower in the control group than in the no AT group. Cardiac troponin I and creatinine were similar among groups.
Comment Antithrombin is a serine protease inhibitor synthesized by the liver. Together with protein C and protein S, it exerts a role as an endogenous anticoagulant. Because
Fig 1. Antithrombin (AT) levels in control patients (Controls; black bars), patients receiving AT (white bars), and those receiving placebo (no AT; gray bars), measured preoperatively (Pre-op), on intensive care unit acceptance (ICU), 90 minutes after the first AT administration (900 AT), 1 hour after the last AT administration (1h AT), 24 hours after the last AT administration (24h AT), 48 hours after the last AT administration (48h AT), and on the fifth postoperative day (day 5). Values are reported as mean standard error.
of catalysis by heparin, its activity is enhanced. Patients undergoing cardiac surgery experience a significant decrease of AT values compared with preoperative levels [4–7], and lower postoperative AT levels have a negative prognostic influence, being independent predictors of several postoperative complications: prolonged ventilation time, low cardiac output syndrome, atrial fibrillation, increased incidence of reintervention for bleeding, need of blood product transfusions, neurological
PAPARELLA ET AL ANTITHROMBIN AFTER CARDIAC SURGERY
1211
complications, and ICU and overall length of stay [5–7]. Currently, AT administration during cardiac operations, especially before CPB establishment in case of heparin resistance, has become a well-established practice [8]. Recently, Ranucci and colleagues [9], in a prospective randomized trial, evaluated the effects of preoperative AT administration to reach an AT activity value of 120% in patients undergoing cardiac surgery with CPB. They showed that preoperative AT administration prevents heparin resistance before CPB establishment and favors higher postoperative AT values, but no clinical benefit could be demonstrated. Postoperative AT administration with the aim of restoring physiologic values is still a therapeutic strategy whose biologic effects and potential clinical benefit have not been documented. Guidelines place AT supplementation for the prevention of thromboembolic complications in selected patient populations in class IIB [10]. For this reason we investigated for the first time the effects of postoperative AT administration in patients with low AT levels after cardiac surgery. Moreover, we analyzed patients with higher AT levels after surgery who served as control subjects to be compared with those undergoing randomization. As previously demonstrated [6], low postoperative AT levels were significantly influenced by preoperative AT levels; in fact, randomized patients had significantly lower AT preoperative values compared with control patients. The cutoff value used to select patients to undergo randomization (
