Recombinant Human Transgenic Antithrombin in Cardiac Surgery: A Dose-finding Study

Patients were enrolled after approval from the Institutional Human Studies Committee (Emory University Hospital, Atlanta, GA) and written informed consent was obtained. Thirty-six patients who were scheduled to undergo elective, primary coronary artery bypass grafting and who had received heparin for at least 12 h before surgery were eligible for this study. Exclusion criteria were previous coronary artery bypass grafting or valve surgery; unstable angina; preexisting hemostatic disorder, renal disease, or hepatic disease; recent or concurrent treatment with warfarin, streptokinase, tissue plasminogen activator, or abciximab (ReoPro^®; Centocor Inc., Malvern, PA); preoperative blood or blood component transfusion; recent or active history of drug or alcohol abuse; known human immunodeficiency virus, hepatitis B, or hepatitis C infections; or known allergy or sensitivity to goat proteins.

This study was of a single-dose, open-label, dose-escalation design. Each of the following dosing cohorts of recombinant human (rh) AT consisted of three patients: 10, 25, 50, 75, 100, 125, 175, or 200 U/kg, respectively. The placebo and 150-U/kg cohorts each consisted of six patients. Patients were enrolled in the next higher-dose cohort after establishment of safety of the previous dose, which was assessed at least 48 h after the last patient within the previous cohort had completed therapy. The decision to advance to the next cohort was reviewed by a safety board composed of the principal investigator from the study site (or a designated representative) and a Medical Affairs representative from the study sponsor (Genzyme Transgenics Corporation, Boston, MA).

Patients’ medical history was recorded and a physical examination was performed within 30 days before surgery. Study medication was reconstituted from a lyophilized vial preparation of rh-AT (7 U/mg specific activity) provided by Genzyme Transgenics Corp. After induction of general anesthesia and tracheal intubation, patients in each dosing cohort received continuous intravenous infusion of rh-AT over 30 min. Patients then received an initial dose of 300 U/kg of porcine intestinal, unfractionated heparin. During CPB, whole heparin concentrations and the activated clotting times (ACTs) were determined at 30-min intervals using a heparin–protamine titration system (Medtronic Perfusion Systems, Minneapolis, MN). Additional heparin was administered as needed to maintain whole blood heparin concentrations between 2.7 and 3.4 U/ml (there were no target ACTs). Coronary artery bypass grafting surgery was performed during moderate hypothermia to 32°C using either vein or artery grafts. After discontinuation of CPB, heparin was neutralized with protamine at a dose ratio of 1.3 mg protamine per milligram of heparin in plasma. The protamine dose was calculated using whole blood heparin concentration measurements performed at the end of CPB. Subsequently, a heparinase-ACT was measured every 2 h after protamine administration for a total of 8 h to detect possible heparin rebound.

Blood specimens were drawn from an indwelling arterial catheter at the following time points: before and 1 min after completion of rh-AT administration; 5 min after heparin bolus administration; 30, 60, and 90 min after initiation of CPB; before protamine administration; and 10 min, 2, 4, 6, 12, and 24 h after protamine administration. AT activity (Coatest Antithrombin; Pharmacia Hepar Inc., Franklin, OH) was measured in all blood specimens. Heparin concentrations (Stachrom Heparin; American Bioproducts, Parsippany, NJ) were measured before rh-AT, after heparin, during CPB every 30 min, and 6 h after protamine. Turbimetric platelet aggregation (platelet aggregometer; Chronolog Corp., Havertown, PA), using 2.4 μm adenosine diphosphate and 2.4 μg/ml collagen (Chronolog Corp.), and bleeding times were measured before rh-AT and 2 h after protamine. The following analyses were performed on all perisurgical samples and at 10 min and 2 and 6 h after protamine: celite ACT using the Hemochron system (International Technidyne, Edison, NJ), kaolin ACT with or without heparinase using the ACT II instrument (Medtronic Perfusion Systems, Minneapolis, MN), heparin concentration (Stockroom Heparin; American Bioproducts), prothrombin fragment 1.2 (Enzygnost F1+2 micro; Behring Diagnostics Inc., Westwood, MA), fibrin monomer (Coatest Fibrin Monomer; Pharmacia Hepar Inc.), D-dimer (Asserachrom D-Di; American Bioproducts), plasmin:antiplasmin complexes, (Enzygnost PAP micro; Behring Diagnostics Inc.), and platelet function and β-thromboglobulin (Asserachrom β-TG; American Bioproducts). Antibody titers were measured before rh-AT administration and again at least 4 weeks after surgery using a qualified indirect enzyme-linked immunosorbent assay developed by Genzyme's Immunology Department and confirmatory radioimmunoprecipitation assay. Based on statistical analysis of the distribution of 200 normal human serum samples, 98% of the population has a final enzyme-linked immunosorbent assay optical density reading less than 0.100. Samples were classified as within normal range or above normal range based on their optical density readings. All serum samples with optical density greater than 0.100 were considered above normal range. A patient with a serum sample within normal range before treatment and above normal range after treatment was considered seroconverted if the above normal range was confirmed as a positive reaction with the confirmatory assay (radioimmunoprecipitation assay).

Hemodynamic variables (e.g. , systolic and diastolic blood pressure, pulse) were monitored before and after AT administration and 24 h after surgery. A postoperative clinical evaluation was obtained at least 1 month after surgery, during which a blood sample was obtained for rh-AT antibody screen. A 12-lead electrocardiograph was obtained before AT treatment (or before heparin in controls) and within 24–48 after surgery. Chest tube drainage was recorded after protamine administration at 10 min and at 2, 4, 6, 12, and 24 h after arrival in the intensive care unit.

One-way analysis of variance was used to compare continuous variables between treatment cohorts, while the chi-square test was used to compare categoric variables between cohorts. Univariate linear regression was used to evaluate the potential linear relation between any two variables. A P  value ≤ 0.05 was considered significant.

The demographic and operative profiles of patients receiving placebo or rh-AT are compared in table 1. Overall, no clinically important safety concerns were discovered in this study. No differences in hemodynamic variables were found between patients receiving rh-AT or placebo.

There were no dose-related trends in chest tube drainage or transfusion requirements (table 2). There was no difference (P = 0.15) for total donor exposures between the placebo (1.08 ± 1.6 U) versus  rh-AT–treated patients (4.8 ± 4.8 U). In addition, there was less transfusion support for the rh-AT-treated patients intraoperatively (1.7 ± 1.8 U) versus  postoperatively (3.1 ± 4.3 U), and there was no significant difference (P = 0.71) in the cumulative blood loss between the placebo-treated (822 ± 329 ml) and rh-AT-treated (893 ± 440 ml) patients.

Table 3summarizes the adverse event frequency in the 30 patients treated with rh-AT and 6 patients in the placebo group. No serious adverse events related to rh-AT administration were reported. One rh-AT–treated patient died. This patient was noted at surgery to have severe diffuse atheromatous disease with loose plaques and suffered an atheromatous embolic stroke during surgery.

Enzyme-linked immunosorbent assay testing for circulating antibodies to rh-AT was positive in three patients: one patient in each 100-, 125-, and 150-U/kg dose groups. Because no antibodies were detected in any patients in higher dose groups (175 and 200 U/kg), there was no relation between dose and reactivity. Positivity (three patients) of enzyme-linked immunosorbent assay testing was not confirmed on radioimmunoprecipitation assay (negative radioimmunoprecipitation test); thus, none of the three patients seroconverted. One of the three patients showed reactivity (above normal range) before receiving rh-AT.

Figure 1shows the dose-dependent increase in plasma antithrombin activity after administration of each dose of rh-AT. With doses that exceeded 50 U/kg, AT plasma concentrations were maintained at greater than 100 U/dl throughout the CPB period. Furthermore, AT concentrations reached 600 U/dl and were maintained at approximately 250 U/dl throughout bypass after 200 U/kg rh-AT, the highest dose administered.

The relation between fibrin monomer and rh-AT dose during and after CPB (n = 226) is shown in figure 2. A significant (P < 0.0001) inverse relation was observed between fibrin monomer concentrations and either rh-AT dose (fibrin monomer =−0.22 AT dose + 34; r =−0.51) or AT concentration (fibrin monomer =−0.25 AT + 44; r =−0.50) when rh-AT dose was less than 150 U/kg. The relation between AT activity and either mean D-dimer values for each dosing cohort or D-dimer concentrations during CPB (n = 121) is depicted in figure 3. A significant (P = 0.001) inverse relation was observed between D-dimer concentrations and either AT activity (D-dimer =−5.8 AT + 1,757; r =−0.49;fig. 3, top) or rh-AT dose (D-dimer =−6.5 AT + 1,586, r =−0.51;fig. 3, bottom). There were no significant differences in prothrombin fragment 1.2 concentrations between placebo and rh-AT groups at any time points studied (P = 0.67, r = 0.02;P = 0.14, r = 0.06); concentrations of prothrombin fragment 1.2 increased during CPB and were still elevated 6 h after administration of protamine. In addition, there were no significant differences between placebo and AT cohorts in concentrations of plasmin–antiplasmin (P = 0.33, r = 0.029;P = 0.74, r = 0.02) and β-thromboglobulin (P = 0.33, r = 0.03;P = 0.74, r = 0.02) concentrations measured over the course of CPB, nor at the preprotamine time point.

In all patients, platelet aggregation values after exposure to adenosine diphosphate and collagen decreased subsequent to protamine administration from the aggregation values observed at baseline. There was apparent effect of supplemental rh-AT (dose) on adenosine diphosphate–induced platelet aggregation (P = 0.006, r = 0.47) at the 2-h postprotamine time point, but no effect was seen in collagen-induced platelet aggregation.

There was no significant difference in bleeding time between placebo and rh-AT patients with respect to rh-AT concentration (P = 0.62, r = 0.12) or AT dose (P = 0.66, r = 0.15). In patients enrolled in the rh-AT dosing cohorts, the increase in median bleeding times from baseline to 2 h after protamine ranged from 0 to 8 min; in patients who received placebo, there was an approximately 4-min increase in the median bleeding time observed after protamine compared with baseline.

Increased kaolin ACT values (P < 0.0001) were observed after systemic heparinization (i.e. , specimen collection time points 3–8) in specimens (n = 134) obtained from rh-AT–supplemented patients (844 ± 191 s) compared with ACT values obtained using specimens (n = 24) from placebo-treated patients (531 ± 180 s). This effect of rh-AT on ACT values appears to be independent of heparin because heparin concentrations (anti-Xa) measured in rh-AT–supplemented patients (5.0 ± 1.5 U/ml) were similar (P = 0.14) to those in placebo-treated patients (4.5 ± 1.6 U/ml, n = 24), and the effect of rh-AT on ACT values persisted even when heparin concentrations were included as a covariate in the analysis during CPB. Anti-Xa measurements at baseline, 5 min after heparin, and at CPB were comparable between the placebo and rh-AT groups: 0.38 versus  0.23 IU/ml (range, 0.15–0.41 IU/ml), 6.43 versus  6.67 IU/ml (range, 5.89–9.06 IU/ml), and 3.89 versus  5.07 IU/ml (range, 3.83–5.32 IU/ml), respectively. Likewise, heparin concentrations were 3.4 (placebo group) versus  4.0 IU/ml (rh-AT group) at 5 min after heparin and 2.7 versus  2.98 IU/ml during CPB.

In this first study of a transgenic recombinant human protein in patients, rh-AT was shown to increase plasma AT concentrations without any serious adverse effects even when supraphysiologic doses were administered to supplement native AT in cardiac surgical patients. Furthermore, administration of rh-AT to 30 patients enrolled in current study resulted in no apparent differences in hemodynamic parameters, chest tube drainage, or transfusion requirements as compared with placebo-treated cohorts (table 2). Because pasteurized AT concentrates have been limited by availability, a new transgenic AT has been developed by modification of goat genome through transgenic technology.

Recombinant human AT is produced in milk of transgenic female goats during parturition. 33,34The initial step in the development of a transgenic goat was creation of an expression vector that contained the sequence for the desired protein linked to the promoter sequence of a specific milk protein gene. This promoter sequence directs the production of the recombinant protein in the goat mammary gland on lactation. Specifically, the human AT cDNA coding for human AT was linked to the caprine β casein gene promoter. 35The β casein gene is regulated in both a tissue and temporal fashion with maximal mammary gland expression occurring after parturition. The goat β casein gene was cloned from a Saanen goat genomic library and characterized in transgenic mice. 36The caprine β casein gene was cloned as an 18.5-kilobase (kb) fragment in a λ EMBL3 vector. The human AT cDNA was obtained in plasmid pβAT6. 37The sequence of cDNA is the same as that published by Bock et al. , 38with the exception of the silent nucleotide changes at base pair 1050 (T-C), 1317 (C-T), and 1371 (A-G). The cDNA was engineered with an Xhol site at the 5′ end by site-directed mutagenesis to allow excision of the cDNA as a 1.45-kb Xhol to Sa1l fragment. 35,36,38–40The 6.2-kb β casein gene promoter was fused to human AT cDNA and 7.1-kb 3′ flanking region of the β casein gene was added to the 3′ end of the AT cDNA to help to stabilize expression levels. 33,35,38,39The 14.75-kb transgene was excited from bacterial sequences and microinjected into preimplantation goat embryos, which were then transferred to the oviducts of surrogate mothers and carried to term. 33,34A founder transgenic goat, a male, was identified for the presence of the transgene by analyzing genomic DNA from both a sample of ear tissue and blood by polymerase chain reaction 41and Southern blot analysis. 38This founder goat was bred to nontransgenic females and produced transgenic progeny. Once established in the first generation of transgenic animals, the transgene is transmitted as other genetic traits to future generations through traditional breeding, thus generating a production herd that supplies the raw rh-AT ready for purification. 34,42A purification process that involves isolation of rh-AT from goat milk has been described with a yield greater than 50% and purity greater than 99.99%. 34,42It involves clarification through a 500-kd tangential flow membrane filtration unit, heparin affinity, anion exchange, and hydrophobic interaction chromatography. This process removes endogenous goat AT present in the milk and other contaminating proteins. 42Structurally, the recombinant human AT purified from goat milk is indistinguishable from plasma-derived AT with the exception of the glycosylation and carbohydrates. 34,42Recombinant AT shows fourfold higher affinity for heparin than plasma-derived AT, which is attributed to the difference in glycosylation between recombinant and plasma-derived AT. 42Recombinant and plasma-derived AT were shown to have equivalent activity in factor Xa and in vitro  thrombin inhibition assays. 42Lu et al.  43investigated pharmacokinetics of a single intravenous dose of recombinant transgenic AT in healthy volunteers (table 4). The concentrations of AT after initial doses (50–200 U/kg) given intravenously over 30 min were best described by a weight-normalized two-compartment model. 43Additional information regarding transgenically produced human AT has been previously reported. 44–46 

Although hereditary AT deficiency is rare, 1acquired AT deficiency is common after noncardiac 47and cardiac surgery. 2–7,48Low AT concentrations during cardiac surgery are likely to develop because of preoperative use of heparin, 3–5the effects of hemodilution, 3–5and CPB-associated excessive hemostatic system activation. 6AT concentrations commonly decrease to 40–50 U/dl activity 32during CPB. 2–7Native AT concentrations as low as 20–30 U/dl were observed during CPB in the current study in patients not receiving supplemental AT. Supraphysiologic AT concentrations were achieved at higher doses of rh-AT, reaching 600 U/dl immediately after the initial loading dose. Concentrations then declined secondary to hemodilution that occurs during initiation of CPB. Doses of rh-AT 75 U/kg normalized AT activity to greater than 100 U/dl during CPB, and AT concentrations remained elevated close to 100 U/dl on arrival in the intensive care unit in patients who received higher doses (i.e. , 150 U/kg). Thus, a single dose of AT maintains normal AT concentrations to cover the clinically important period of CPB during which excessive hemostatic system activation may occur and can lead to bleeding or thrombotic complications.

Anticoagulation is used during cardiac surgery to prevent thrombosis of the extracorporeal circuit and to minimize excessive CPB-related activation of the hemostatic system. Unfractionated heparin is routinely used because it is immediately effective, rapidly reversible, generally well tolerated, and inexpensive. 49The heparin anticoagulant effect is predominately mediated by binding of heparin to AT, requiring a specific pentasaccharide sequence, 50and subsequent irreversible complex formation and inactivation by AT of several activated coagulation factors, in particular, factors IIa and Xa. 51Decreased heparin anticoagulant response or heparin resistance is associated with AT deficiency. 52This has led to the use of either fresh frozen plasma 53or human plasma-derived AT concentrates (AT concentrates are not approved for this indication in the United States) 5,54in patients that show appreciable resistance to heparin before initiation of CPB. 53Administration of rh-AT in the current study improved heparin anticoagulant effect, as reflected by increased ACTs during heparin therapy when compared with placebo patients.

Detection and treatment of acquired AT deficiency may also be important with respect to potential reduction in perioperative bleeding and thromboembolic complications. CPB-related activation of the hemostatic system 28–30can consume labile coagulation factors and platelets, which may lead to excessive bleeding (i.e. , in 5–16% of patients) 10,11or reexploration (3–4%) 11,12after cardiac surgery. Acquired AT deficiency may render heparin ineffective in suppressing thrombin generation–activity during CPB. 3,5,8This is important because recent evidence indicates that inadequate heparin anticoagulation can result in excessive bleeding and transfusion requirements 55secondary to consumption of coagulation factors and platelets. 6Therefore, detection and treatment of AT deficiency may be important with respect to preservation of labile coagulation factors and platelets, which may reduce blood loss and need for transfusion. The importance of AT in suppressing activation of the hemostatic system is supported by data from several studies. 3,52Hashimoto et al.  3showed that AT supplementation more effectively suppresses thrombin activity in pediatric patients who received AT concentrates. Despotis et al.  52demonstrated that appreciable increases in markers of platelet activation and coagulation activation were only seen when AT concentration decreased to less than 60 U/dl.

Because maintenance of normal or elevated plasma AT concentrations during CPB can enhance suppression of the hemostatic system, the current study investigated the efficacy of increasing doses of transgenic rh-AT to suppress markers of hemostatic activation. The findings of the current study indicate that normalizing AT concentrations with rh-AT during CPB can not only reduce thrombin activity (fibrin monomer), but can also reduce fibrinolytic activity (D-dimer) and reduce impairment of platelet function after CPB, which is thought to be the most important hemostatic defect after CPB. 56,57Although this study did not show any reductions in excessive microvascular or perioperative thrombotic complications in patients receiving rh-AT, this could be because of the small sample size and low-risk population studied. Because we studied a small number of patients in each group, this study was not powered to prevent significant type II error; therefore, additional studies involving more patients will be necessary in the future to fully assess benefits of supplemental administration of rh-AT to cardiac surgical patients. Nevertheless, our data support the concept that rh-AT administration enhances heparin responsiveness and modulates activation of the hemostatic system.

Perioperative thrombotic complications such as deep venous thrombosis, pulmonary embolism, stroke (1.6–5.6%), 23–25and perioperative myocardial infarction–injury (5–82%) 26,27occur with cardiac surgery. Coronary, 17,18,22pulmonary, 19,21and intracardiac 20thrombosis have also been described intraoperatively. Surgery and CPB result in excessive activation of the hemostatic system via  both the intrinsic 28and extrinsic 29,30pathways, which leads to substantial thrombin generation and a prothrombotic state. The vascular endothelium and multiple circulating factors, including endothelium-derived relaxing factor, prostacyclin, tissue plasminogen activator, endothelium heparan sulfate, proteins C and S, and circulating AT normally inhibit intravascular thrombosis. However, normal circulating anticoagulants such as AT and proteins C and S are commonly reduced secondary to CPB-related consumption and hemodilution, 2–7,31which may lead to hypercoagulability in the perioperative cardiac surgical period and thrombotic complications. Because perioperative bleeding and thrombotic complications after cardiac surgery may be related to inadequate suppression of the hemostatic system during and after cardiac surgery, detection and treatment of acquired AT deficiency may have important clinical consequences.

This is the first reported use of a transgenic recombinant protein for therapeutic application. Reductions in AT concentrations that occur during heparin therapy and extracorporeal circulation represent an important clinical application for rh-AT. Our data indicate that administration of transgenic rh-AT to patients and volunteers shows no adverse effects, can improve heparin responsiveness at doses used for CPB, and potentially preserve the hemostatic system via  enhanced anticoagulation during CPB. Because our study included only a small number of patients in each dosing cohort, further studies with larger number of patients are needed to determine the clinical usefulness of rh-AT with respect to management of heparin resistance, prevention of excessive bleeding, and blood conservation, as well as prevention of thromboembolic sequelae. Other potential therapeutic applications that warrant investigation include administration of rh-AT to patients with unstable angina on heparin therapy or in patients undergoing angioplasty or other arterial manipulations.