Within coronary patients, those who are affected by

The Role of Antithrombin III in the Perioperative Management of the Patient With Unstable Angina Marco Rossi, MD, Lorenzo Martinelli, MD, Sergio Storti, MD, Michele Corrado, MD, Roberto Marra, MD Carmelita Varano, MD, and Rocco Schiavello, MD Departments of Anesthesia and Critical Care, and Hematology, Catholic University of the Sacred Heart, Rome, Italy Background. To evaluate the effectiveness of intraoperative administration of antithrombin III (AT III) to improve anticoagulation and preserve the hemostatic mechanisms during cardiopulmonary bypass (CPB) in patients with unstable angina under heparin treatment. Methods. We divided 22 patients, scheduled for coronary artery bypass grafting, into two groups. Group A (11 patients) received 3000 International Units (IU) of AT III concentrates plus heparin before aortic cannulation. Group B (11 patients) received only heparin. Blood drainage, allogeneic blood transfusions, and intraoperative activated coagulation time were recorded. Also, AT III, thrombin-antithrombin complex (TAT), fragment 1.2 (F 1.2), and D-dimers were measured during the operation and the first postoperative day. Results. Group A patients had fewer transfusions and had less chest-tube drainage. In group A, AT III levels increased after AT III concentrates administration and were always higher than in group B. In group B, F 1.2 and TAT increased significantly more after CPB and at the end of operation. Differences in D-dimers between the groups were not significant. Conclusions. Intraoperative administration of AT III concentrates allowed adequate anticoagulation during CPB and attenuated the coagulative cascade activation and the consequent consumptive coagulopathy. (Ann Thorac Surg 1999;68:2231 6) 1999 by The Society of Thoracic Surgeons Within coronary patients, those who are affected by unstable angina represent a growing share; for these patients, surgery is often the only possible choice [1]. Preoperative treatment includes heparin as the main point to prevent angina episodes and angina s most dangerous complications, acute myocardial infarction and sudden death [2 4]. Preoperative heparin therapy causes a consumption of antithrombin III (AT III) with a consequent resistance to heparin due to the deficiency of the cofactor [5 7]. The use of higher doses of heparin in these patients does not always reach the aim of an effective anticoagulation during cardiopulmonary bypass (CPB), which is proved by the inadequate ACT values obtained [8]. In this scenario, the phenomena linked to the consumption coagulopathy are amplified, causing a heavier postoperative bleeding and increased risk of surgical revision of hemostasis and allogeneic blood transfusions [9 10]. Preservation of the physiological levels of AT III in those patients who should undergo CPB seems to be the most rational and effective therapeutic approach [11 14]. The aim of this work is to evaluate the effectiveness of an intraoperative single dose of AT III concentrates in patients with unstable angina and under preoperative heparin treatment in improving anticoagulation and preserving the hemostatic mechanisms. Accepted for publication June 7, 1999. Address reprint requests to Dr Rossi, Istituto di Anestesiologia e Rianimazione, Università Cattolica del Sacro Cuore, Largo A. Gemelli, 8, 00168 Rome, Italy; e-mail: mrossi@edl.it. Material and Methods A total of 22 patients (21 men and 1 woman, mean age 61 years 9) affected by ischemic heart disease and unstable angina (Canadian Cardiovascular Society Grading Scale Class IV) and scheduled for elective coronary artery bypass surgery (CABG) were enrolled into the study after informed consent and approval by the local ethical committee. All patients were receiving subcutaneous heparin ranging from at least 72 hours before operation, to a maximum of 1 week. Patients studied were randomly divided into two groups: the AT III group (group A), of 11 patients and the control group (group B), of 11 patients. Exclusion criteria for the investigation were: (1) age greater than or equal to 70 years, (2) hepatic, renal diseases or congestive heart failure (CHF), (3) patients having oral anticoagulant or intravenous heparin or acetylsalicylic acid treatment, and (4) patients having intravenous nitrates because of the nitrates induced heparin resistance [15]. The anesthetic management, operative procedure, and conduct of CPB were standardized. A hollow-fiber membrane oxygenator (Monolyth, Sorin Italy) was used in all cases; CPB was always normothermic. The same surgical team performed all the operations and were not aware of the group allocation of the patients. Before CPB, AT III concentrate, 3000 International Units (IU) (Kybernin Behring, Marburg, Germany), was administered to the patients of group A in addition to the heparin, 300 U/kg (Eparina Bristol-Myers Squibb, Anagni, 1999 by The Society of Thoracic Surgeons 0003-4975/99/$20.00 Published by Elsevier Science Inc PII S0003-4975(99)00864-4

2232 ROSSI ET AL Ann Thorac Surg ANTITHROMBIN III IN CORONARY SURGERY 1999;68:2231 6 Italy). Only heparin (300 U/kg) was administered to the patients of the control group. An additional 10000 units of heparin were included in the bypass circuit of all patients. Celite-based activated coagulation time (ACT) was maintained at 480 seconds or greater in all patients by the addition of heparin when necessary. Heparin was neutralized by an intravenous injection of protamine sulfate (Protamina Roche, Milan, Italy) within 5 minutes after the end of CPB, in a ratio of 1.3 mg of protamine for every 100 units of total heparin administration. The following data were recorded: Duration of the operation, CPB, and aortic crossclamping Time taken between the end of CPB and the closure of the operative site Loss of fluid into the wound drainage system in the first 24 postoperative hours Amount of residual volume in the CPB circuit collected in blood transfusion bags and retransfused Amount of autologous blood withdrawn before starting CPB and retransfused after protamine administration Amount of allogeneic blood or blood products required. The indication for transfusion was defined as a hematocrit less than 22% during CPB and less than 30% in the postoperative period Use of antifibrinolytics Amount of heparin and protamine administered Surgical reopening for postoperative bleeding ACT was recorded at the following times: before induction, after administration of heparin, after the start of CPB, at the end of CPB, after administration of protamine, at the end of operation, first postoperative hour in intensive care unit (ICU). Blood samples were collected for measurement of prothrombin time (PT), activated partial thromboplastin time (aptt), and fibrinogen and platelets at the following times: before operation, end of operation, and first postoperative day. Whole blood samples were collected from a nonheparinized line of a triple-lumen central venous catheter to measure: AT III Thrombin-antithrombin complexes (TAT) Fragment 1.2 of prothrombin (F 1.2) Split products of the cross-linked fibrin (d-dimers) at the following times: before induction; after AT III concentrates administration and before start of CPB; end of CPB; end of operation; and first postoperative day. We measured AT III as heparin cofactor with a functional assay, based on the inhibition of thrombin-induced proteolysis of a specific chromogenic substrate (polypeptide-pna conjugated) (Antitrombina III, Instrumentation Laboratory Co, Lexington, KY). We measured AT III as a percentage, and reference values ranged between 70% and 120%. We measured TAT with an immunoenzymatic Table 1. Patients Characteristics photometric test (Enzyguest-TAT micro Behring, Berhingwerke AG, Marburg, Germany), with results expressed in g/ml. We measured F 1.2 with the enzyme-linked immunosorbent (ELISA) method (Thrombonostika F 1.2 sample, Organon Teknica Corp, Durham, NC) using specific monoclonal antibodies. Reference value was 1 nmol/ml. Semiquantitative determination of d-dimers was obtained using a latex agglutination test (Fibronosticon, Organon Teknica Corp, BV, Boxtel, The Netherlands). We measured d-dimers in ng/ml and a value greater than 2000 ng/ml was considered as significant. All values were expressed as mean and standard deviation. Comparisons among dichotomous variables were performed by 2 test, and Student s t test was used to compare continuous variables. The comparison between the groups was performed by two-factor ANOVA for repeated measures. ANOVA p values for intragroup (factor time) and intergroup variations (factor treatment) were reported. In case of significant differences between the two groups, the analysis was completed with Newman-Keuls test, and p values for the comparison at single times were reported. Values of p less than 0.05 were considered as significant. Results Group A: Treated with AT III Group B: Control Patients 11 11 Mean age ( SD) 61 9 60 10 Patients with LIMA 9 11 CABG 1 1 0 CABG 2 6 6 CABG 3 3 5 CABG 5 1 0 Distal anastomoses (mean) 2.45 2.45 ATIII antithrombin III; CABG coronary artery bypass graft; LIMA left internal mammary artery. Patients characteristics with respect to age, number, and type of grafts were similar in the two groups (Table 1). Times of the operation were not significantly different in the two groups of patients (Table 2). The amount of the autologous blood withdrawn before CPB and retransfused after protamine was similar in the two groups of patients (659 102 ml in group A; 650 71 ml in group B). A similar amount of residual volume collected from the CPB circuit was retransfused in the two groups after surgery during the first 24 postoperative hours (1,632 444 ml in AT III group and 1,609 347 ml in the untreated group). None of the 22 patients needed reexploration for surgical bleeding. Only 1 patient in the group treated with AT III and 4 patients (36%) in the control group received allogeneic blood perioperatively. No patients in either group required transfusion of platelets or fresh frozen plasma.

Ann Thorac Surg ROSSI ET AL 1999;68:2231 6 ANTITHROMBIN III IN CORONARY SURGERY 2233 Table 2. Comparison of Operating Parameters Group A Group B t Test Operation time (min) 226 34 238 41 ns Clamping time (min) 54 19 58 13 ns CPB time (min) 65 20 68 13 ns Chest closure time (min) 67 16 60 24 ns Bleeding (ml/24 h) 522 182 706 127 p 0.05 Heparin (mg) 339 58 510 62 p 0.05 Protamine (mg) 450 86 618 106 p 0.05 Surgical re-exploration 0 patients 0 patients 2 ns Transfusion 1 patient 4 patients 2 ns Table 4. Hematocrit Variations HT % Group A Group B Newman-Keuls Group A B Baseline 42 4 a 42 4 a ns Before CPB 33 6 a 30 5 a ns End CPB 26 3 a 26 3 a ns End operation 30 4 a 32 3 a ns 1st postoperative day 34 4 a 35 4 a ns a ANOVA factor time within group: p 0.001. HT hematocrit; CPB cardiopulmonary bypass; ns not significant. CPB cardiopulmonary bypass; ns not significant. The chest-tube drainage in the first 24 postoperative hours was significantly greater in the control group than in the group treated with AT III (706 127 ml versus 522 182 ml, p 0.05). Heparin intraoperative requirement was significantly lower in the patients treated with AT III than in the patients of the control group (339 58 mg versus 510 62 mg, p 0.05) (Table 2). Whereas in group A there was no need for additional heparin doses, in 7 patients of group B (67%) additional heparin doses were required to keep ACT values above 480 seconds. After heparin and during CPB, ACT values were significantly higher in the treated group than in the untreated group (Table 3). The total amount of protamine administered to the patients of the AT III group was significantly lower than that administered to the patients of group B (480 86 mg versus 618 106 mg, p 0.05) (Table 2). Hematocrit values changed significantly at the scheduled times in each group, but the differences between the two groups were not statistically significant (Table 4). In both groups, PT, aptt, and fibrinogen values varied significantly but similarly at the end of operation and at the first postoperative day (Table 5). Platelets significantly decreased at the end of the operation in both groups (172,000 40,000 versus 216,000 59,000 in group A; 142,000 35,000 versus 210,000 56,000 in group B) and returned to the preoperative values at the first postoperative day in both groups without statistically Table 3. ACT Variations ACT(s) Group A Group B Newman-Keuls Group A B Baseline 200 31 a 219 37 a ns Before CPB 777 69 a 537 81 a p 0.05 Start CPB 919 280 a 559 160 a p 0.01 End CPB 740 176 a 525 175 a p 0.01 After protamine 153 29 a 168 41 a ns End operation 143 33 a 151 25 a ns 1st h in ICU 136 28 a 152 17 a ns a ANOVA factor time within group: p 0.01. ACT Activated coagulation time; CPB cardiopulmonary bypass; ICU intensive care unit; ns not significant. significant differences between the two groups at the three scheduled times (Table 5). The data about the hemocoagulative markers discussed in the present study are reported in Figures 1 and 2. Preoperative AT III levels were at the lower limits of the normal range in both groups (75.3% 4.1 in group A; 76.1% 7.5 in group B). No significant change in AT III was shown at the scheduled times in the control group except for a small decrease at the end of CPB (68.1% 7.4). On the contrary, AT III increased significantly in the treated group compared to the preoperative value (p 0.001). In group A, AT III values showed a sharp rise (102.2% 30.1) after the administration of AT III concentrates before entering CPB, which was maintained during the following times except for a small decrease at the end of CPB (90.5% 18.3). In the treated group, AT III values, except for the preoperative value, were always significantly higher than AT III values of the untreated group (p 0.01). In both groups, TAT, F 1.2 and d-dimers showed important changes at the different times of the study (p 0.001). In fact, all three markers showed a sharp rise at the end of CPB and these levels were maintained at the end of the operation. Levels of TAT increased significantly more in the control group than in the AT III group at the end of CPB (60.2 1 g/l versus 28.4 15 g/l, p 0.01) and at the end of operation (59.4 3 g/l versus 30.8 5 g/l, p 0.01). A similar course was seen with F 1.2, which increased more in the untreated patients at the end of CPB (2.7 0.6 nm/ml versus 1.7 0.3 nm/ml, p 0.01) and at the end of operation (2.2 0.4 nm/ml versus 1.7 0.3 nm/ml, p 0.01). Variations in d-dimers were not significantly different between treated and untreated patients at the end of CPB (420.8 175 ng/ml versus 510.3 157 ng/ml) and at the end of the operation (377.5 165 ng/ml versus 480.9 167 ng/ml), even if they increased more in the untreated patients. Whereas TAT and d-dimers levels remained significantly higher than the preoperative values (p 0.05) in both groups at the first postoperative day, F 1.2 returned to the preoperative values. At the first postoperative day, TAT, d-dimers and F 1.2 levels did not differ significantly between the groups.

2234 ROSSI ET AL Ann Thorac Surg ANTITHROMBIN III IN CORONARY SURGERY 1999;68:2231 6 Table 5. Coagulation Data Group A Group B Newman-Keuls Group A B Platelets mm 3 Baseline 216 59 b 210 56 b ns End of operation 172 40 b 142 35 b ns 1st postoperative day 184 38 b 164 43 b ns PT % Baseline 102 12 b 96 10 b ns End of operation 68 6 b 65 8 b ns 1st postoperative day 72 9 b 74 5 b ns aptt second Baseline 29 8 a 32 2 b ns End of operation 29 9 a 43 5 a ns 1st postoperative day 53 11 a 47 7 a ns Fibrinogen mg/dl Baseline 324 45 b 343 83 b ns End of operation 275 123 b 254 45 b ns 1st postoperative day 520 142 b 498 91 b ns a ANOVA factor time within group: p 0.01. b ANOVA factor time within group: p 0.001. aptt activated partial thromboplastin time; PT prothrombin time. Comment The use of CABG operations is increasing due to the progressive spreading of atherosclerotic coronary disease and to a better diagnosis of the patients at risk. The patient with unstable angina on preoperative heparin treatment might come to the operation with a poor coagulative system, because a prolonged heparin treatment causes a decrease of AT III levels. Cofactor deficit might lead to an inadequate systemic anticoagulation during CPB and to the phenomena of consumptive coagulopathy due to excessive thrombin activation [5, 6]. This study showed the effectiveness of an intraoper- Fig 1. AT III (A) and TAT (B) comparison between the groups at the times of the study. (**Newman-Keuls A B p 0.01. AT III antithrombin III; TAT thrombin-antithrombin complex.) Fig 2. F 1.2 (A) and d-dimers (B) comparison between the groups at the times of the study. (**Newman-Keuls A B p 0.01. F 1.2 fragment 1.2 of prothrombin.)

Ann Thorac Surg ROSSI ET AL 1999;68:2231 6 ANTITHROMBIN III IN CORONARY SURGERY 2235 Fig 3. The scheme shows AT III sites of action in the final pathway of the coagulation system, localizing TAT, F1.2, and d-dimers within the cascade. (AT III antithrombin III; F 1.2 fragment 1.2 of prothrombin; TAT thrombin-antithrombin complex.) ative AT III concentrates treatment in patients with unstable angina undergoing coronary artery bypass surgery. The group of patients treated with AT III concentrates showed a better intraoperative and postoperative coagulative pattern and expression of a more correct functioning of the hemostatic balance. In fact, in all the treated patients, standard heparinization proved to be effective to keep the optimal anticoagulation during CPB, as confirmed by ACT values constantly above 480 seconds. Both groups have AT III preoperative levels in the inferior part of the normal range because of the prolonged preoperative heparin treatment. The administration of AT III concentrates to group A restores adequate levels of AT III, allowing heparin to be the antagonist of thrombin and to stop the coagulative processes triggered by CPB. In group B, heparin results were not as effective, because AT III preoperative levels were not sufficient to obtain a complete anticoagulation during CPB. So, to keep adequate ACT, extra doses of heparin are required. If lower concentrations of AT III reduce heparin effect, higher levels of free thrombin are produced, even if more heparin is empirically administered. Free thrombin is able to activate a consumptive coagulopathy. The behavior of TAT, F 1.2, and d-dimers confirms this hypothesis. These factors are strictly linked to thrombinantithrombin III interaction as shown in Figure 3. In the treated group, F 1.2 increases significantly less after CPB, so less thrombin has been generated; this suggests that AT III concentrates should be able to reestablish a better hemostatic balance, playing a protective role against consumptive coagulopathy, even if consumptive coagulopathy is not completely prevented. The TAT increase in the control group at the end of CPB and at the end of the operation suggests that more free thrombin was produced in this group during CPB, which somehow consumes more available AT III. Though not significantly, d-dimers increased more in the untreated group after CPB. This shows that a higher level of free thrombin was generated, which was able to act on fibrinogen. Heparin worked better in group A, as confirmed by the fact that the treated patients did not need any extra heparin dose to keep ACT above 480 seconds. If heparin administration during the operation was increased, it was possible to obtain ACT above 480 seconds in the untreated patients too. This was probably due to the antithrombinic action of heparin through the heparinic cofactor II. However, if more heparin is given, there are more possibilities to find it bonded to the endothelium or in the residual volume in the CPB circuit reinfused after the operation. This might cause an increased risk of heparin rebound and motivate a trend to administer higher doses of protamine, based on the clinical observation of a wet patient. In fact, more protamine was administered to, and more blood loss and transfusion requirements were observed in, the untreated group. Protamine itself can act as anticoagulant, inhibiting the thrombin, when administered in excessive doses or in the absence of heparin. In our institution, the majority of patients affected by unstable angina are treated with subcutaneous heparin, beta blockers, and nitrates po. We excluded patients treated with intravenous nitroglycerin because of the possible induction of heparin resistance. The therapy with intravenous heparin and intravenous nitrates is limited to severe unstable angina, with indication for urgent revascularization. This study shows clearly that intraoperative administration of AT III concentrates in patients with unstable angina having preoperative heparin treatment allows an adequate anticoagulation during CPB and avoids, though incompletely, the activation of the coagulative cascade. The high cost of AT III concentrates should be considered, but it should be compared with the costs of an allogeneic blood transfusion or surgical reexploration. The risk of virally transmitted diseases can not be completely excluded, although it is very low, because of the safe techniques of purification in selected donor pools. Patients with unstable angina having preoperative heparin treatment undergoing coronary surgery in CPB should be always considered as patients at higher risk of bleeding, and the surgical team must be aware of their possible coagulative derangement. References 1. Naunheim KS, Fiore AC, Arango DC, et al. Coronary artery bypass grafting for unstable angina pectoris: risk analysis. Ann Thorac Surg 1986;47:569 74. 2. Collins P, Fox KM. Pathophysiology of angina. Lancet 1990;1:94 110. 3. Neri Serneri GG, Gensini GF, Poggesi L, et al. Effect of heparin or aspirin in reduction of myocardial ischemia in refractory unstable angina. Lancet 1990;335:615 8. 4. Theroux P, Ouimet H, McCans J, et al. Aspirin, heparin or both to treat unstable angina. N Engl J Med 1989; 319:1105 11. 5. Lidon RM, Theroux P, Robitaille D. Antithrombin III plasma activity during and after prolonged use of heparin in unstable angina. Thromb Res 1993;72:23 32.

2236 ROSSI ET AL Ann Thorac Surg ANTITHROMBIN III IN CORONARY SURGERY 1999;68:2231 6 6. Dietrich W, Spannagl M, Schramm W, et al. The influence of preoperative anticoagulation heparin response during cardiopulmonary bypass. J Thorac Cardiovasc Surg 1991; 102:505 14. 7. Anderson EF. Heparin resistance prior to cardiopulmonary bypass. Anesthesiology 1986;64:504 12. 8. Karski J, Teasdale S, Norman P, et al. Heparin requirements before and during CPB in patients receiving preoperative intravenous heparin. Anesthesiology 1994;77:A152. 9. Soloway HB, Christiansen TW. Heparin anticoagulation during cardiopulmonary bypass in an antithrombin III deficient patient. Am J Clin Pathol 1980;73:723 5. 10. Woodman RC, Harker LA. Bleeding complications associated with cardiopulmonary bypass. Blood 1990;76: 1680 97. 11. Buller H, Cate JW. Acquired antithrombin III deficiency: laboratory diagnosis, incidence, clinical implications and treatment with antithrombin III concentrates. Am J Med 1989;87(Suppl 3B):44S 48S. 12. Clark P, Walker ID, Neilson F. Coronary artery bypass surgery in patients with inherited antithrombin deficiency. Br J Haematol 1995;90:479 82. 13. Lechner K, Kyrle A. Antithrombin III concentrates. Are they clinically useful? Thromb Haemost 1995;3:340 8. 14. Menaché B, Grossman BJ, Jackson CM. Antithrombin III: physiology, deficiency and replacement therapy. Transfusion 1992;2:580 88. 15. Becker CR, Corrao J, Bovill GE, et al. Intravenous nitroglycerin-induced heparin resistance: a qualitative antithrombin III abnormality. Am Heart J 1990;119:2231 6. INVITED COMMENTARY This is an interesting study showing a dramatic decrease in the initial heparin dose requirement as well as in heparin requirements during CPB in patients pretreated with ATIII after subcutaneous heparinization. I am much less impressed with reduction of blood loss shown in this study (average of 184 ml over 24 hours). The small number of patients enrolled in this study makes it impossible to evaluate any clinically relevant benefit of ATIII treatment to these patients. Considering the cost of ATIII and concerns about unanswered questions (use of ATIII in cardiac surgical patients not on heparin, use after prolonged IV heparinization, comparison with antifibrinolytics with respect to bleeding), I would wait for large clinical dose finding studies and outcome trials before recommending routine use of ATIII in cardiac patients. Jacek M. Karski, MD Anaesthesia Dept The Toronto Hospital 200 Elizabeth St Toronto, ON M5G 2C4, Canada 1999 by The Society of Thoracic Surgeons 0003-4975/99/$20.00 Published by Elsevier Science Inc PII S0003-4975(99)01002-4