THROMBOSIS RESEARCH 67; 233-241,1992 0049-3848/92 $5.00 + .OOPrinted in the USA. Copyright (c) 1992 Pergamon Press Ltd. All rights reserved.
RECOMBINANT HUMAN FACTOR VHa (rFVIIa) IN A RABBIT STASIS MODEL
Viggo Diness, Claus Bregengaard, Elisabeth Erhardtsen and Ulla Hedner Novo Nordisk A/S, Novo Alle, DK-2880 Bagsvaerd, Denmark
(Received 105.1992; accepted in revised form 22.6.1992 by Editor H.A. Vinazzer)
ABSTRACT The thrombogenicity of recombinant human FVIIa (rFVIIa) and FEIBA was studied in a rabbit stasis model. Only minor thrombus formation in isolated vein segments during 10 min. of stasis was seen following the administration of rFVIIa 100-1000 pg/kg or FEIBA 50-100 U/kg whereas both compounds caused clear thrombus formation during 30 min. of stasis. RFVIIa caused no change in platelet counts or plasma fibrinogen 3 hours after administration, whereas a small decrease of AP’IT was seen due to the direct effect of rFVIIa in this assay. In contrast, FEIBA caused a significant and dosedependent decrease of platelet counts as well as of fibrinogen, and an increase of APTT which demonstrate a general consumption of coagulation factors. In conclusion the study demonstrated a local pharmacological effect of rFVIIa in the absence of a general activation of the coagulation cascade.
INTRODUCTION It has been demonstrated that recombinant human factor VIIa (rFVIIa) is effective in the treatment of bleeding episodes in haemophilia patients with inhibitors against FVIII or FIX (1,2). Such patients are currently treated with plasma preparations containing activated coagulation factors claimed to by-pass the factor VIII/factor IX-dependent pathway of the coagulation system. This treatment is associated with the well-known thrombogenic risk of such preparations which may lead to disseminated intravascular coagulation (DIC) (3,4).
Key Words: FVIIa, DIC, rabbit stasis model
233
234
RECOMBINANT Vlla
Vol. 67, No. 2
In the present study the potential thrombogenicity of rFVIIa was examined in a rabbit stasis model. These results were compared to results obtained on the plasma-derived preparation FEIBA. Also potential interaction between rFVIIa and FEIBA was examined.
MATERIALS AND METHODS Operation: New Zealand white rabbits, females 2.5-3.3 kg body weight, were used. Anaesthesia was induced by intravenous administration of about 2 ml of pentobarbital sodium 5% via the marginal ear vein and was maintained throughout the experiment by additional administrations of about 0.1 ml per 15 min. according to effect. The rabbit was placed on the back. A PE-90 catheter was placed in a. carotis and kept open by infusion of saline. This catheter was also connected to a pressure transducer (Gould P 23 XL) and recorder (Kipp & Zonen) for continuous registration of blood pressure. A PE-50 catheter was placed in v. femoralis for administration of test compounds. Finally, v. facialis in both sides were isolated and non-occlusive ligatures were placed around the veins. Materials: Recombinant human FVIIa was produced by Novo Nordisk as previously described (5). In the final preparation, rFVIIa is the activated two-chain molecule with a specific activity of > 20,000 U/mg in a one-stage clotting assay (6). The amount of rFVIIa (mg) was determined using a sandwich ELISA based on two monoclonal antibodies (7). The freeze-dried preparation was reconstituted in sterile water to a concentration of 0.6 mg/ml which was used without further dilution. FEIBA (Immuno) (500 units), commercially available, was reconstituted in 20 ml of sterile water according to the recommendations of the manufacturer. These 20 ml were stored at -80°C until use. Immediately before intravenous administration this solution (25 U/ml) was thawed and further diluted in saline + 1% bovine serum albumin to final concentrations of about 15-18 U/ml according to the weight of the rabbit. Administration of test compounds. The experimental groups are shown in Tables l-2. A bolus of saline 1 ml/kg was given as control (group 1). RFVIIa 100, 300, and 1000 pglkg was given as a bolus during 1 min. (groups 2-4). FEIBA 50 and 1Oc’U/kg (groups 5-6) was given as an infusion over a period of 15 min. Group 7 receive? FEIBA 50 U/kg. Five minutes after the termination of FEIBA infusion rFVIIa 100 pglkg was given as a bolus. After administration of each compound the catheter was flushed with saline. The study was carried out randomized and according to Good Laboratory Practice (GLP). Stasis. Ten minutes after administration of the test compound(s) an atraumatic clamp was placed on one facial vein close to the jugular vein, thereby causing total stasis. Blood was allowed to fill the facial vein. Then a second clamp was placed on the vein, thus isolating a segment of about 1.5 cm between the two clamps. One minute later this procedure was repeated on the facial vein at the other side. After 10 minutes of stasis, the loose ligatures at one side were tightened and the vein segment was removed and inspected for the presence of thrombus formation. This procedure was repeated at the other side after 30 minutes
Vol. 67, No. 2
RECOMBINANT Vlla
235
of stasis. If present, thrombi were transferred to pre-weighed vials for determination of wet weight. Blood samples. Blood samples were taken via the catheter in a. carotis administration of the test compound(s) and 5 min. and 3 hours after, respectively.
before
Analyses. Whole blood platelet counts, plasma concentration of fibrinogen and activated partial thromboplastin time (APTT) were determined in all pretreatment and 3-hoursamples. Determination of platelet counts in EDTA-stabilized blood was carried out in an Analysis 148 C hemocounter. Plasma was prepared from blood anticoagulated with sodium citrate (final concentration 0.013 M) and stored at -80°C until analysis. Fibrinogen was determined according to the method described by I.M. Nilsson and B. Olow (8). This method includes coagulation of plasma by thrombin (Topostasin, Roche) and isolation of the fibrin coagulum. The detection limit of fibrinogen was found to be 35% of rabbit reference plasma levels. Activated partial thromboulastin time (AP’IT) was measured in a Behnk Coagulator using CepQotest (Nycomed, Oslo) according to the recommendations of the manufacturer. It was controlled whether rFVIIa or FEIBA interfered directly with the assays. RFVIIa antigen was determined in the 5-min. and 3-hour-samples from groups 2, 3 and 4. Furthermore, 10 ~1 or 30 ~1 of the injected preparation was added to 1 ml of predose plasma which was analyzed in parallel with the postdose samples. All plasma samples for determination of rFVIIa antigen were diluted 1: 10 in the ELISA dilution buffer before storage at -80°C. The ELISA assay (NovoClone, Novo Nordisk) is based on two monoclonal antibodies and does not detect endogenous FVII in the rabbit (7). The dilution buffer is a phosphate buffer, pH 7.2, which contains Tween 20 and bovine serum albumin, 1%. Calculations. Platelets (X 109/1), fibrinogen (g/l), and APTT (sec.) 3 hours after administration of test compounds were all calculated as per cent (%) of the corresponding predose value in each experiment. These relative values were used as the final results and subjected to statistical analysis. Statistics. Mann Whitney’s test was used for all statistical evaluations,
RESULTS Local thromhus formation: Results are given in Table 1. No thrombus formation during 10 or 30 min. of stasis was seen in the saline control group (Table 1). During 10 min. of stasis very little thrombus formation was seen after administration of, rFVIIa lOO1000 pg/kg, a small thrombus being formed in only 3 out of 18 vein segments. The mean thrombus weight was less than 1 mg (total range O-3.6 mg). Also FEIBA 50-100 U/kg induced little thrombus formation during 10 min. of stasis. During prolonged stasis (30 min.) rFVIIa as well as FEIBA caused clear thrombus formation in almost all vein segments. When rFVIIa 100 pglkg was given in conjunction with FEIBA 50 U/kg thrombus formation was more rapid since thrombi of approximately 10 mg were seen already after 10 min. of stasis (p < 0.01 for group 7 versus groups 2, 5 and 6).
236
RECOMBINANT Vlla
Vol. 67, No. 2
TABLE 1 Local thrombus formation in isolated vein segments. Number of thrombi and mean thrombus weight (mg) after 10 and 30 min. of stasis (n = 6) 10 min. stasis Test Group
I 1 haline 2
I
6
rFVIIa 100 pglkg
I FEIBA
100 U/kg
30 min. stasis
I
Number of thrombi
Mean weight (mg)
Number of thrombi
Mean weight (mg)
0
0
0
0
I
1
1
0.4
1
4
1
3.2
1
1
1
1.5
1
6
I
7.1
I
9.8
1
6
(
8.2
1
6
*n=5 Analyses: In the control group the predose values were: platelets 307f54 x 109/1, fibrinogen 2.4kO.3 g/l, APTT 27.4k3.2 set (mean *SE). The postdose values were calculated as per cent of the predose values in all groups (Table 2). In the saline control group platelet counts, fibrinogen, and AP’IT changed little during the experiment, the postdose values being about 90% of the predose values. Compared to the control group rFVIIa 100-1000 pg/kg caused no change of platelet counts or fibrinogen, whereas APTT was slightly shortened. In contrast, FEIBA caused a significant and dosedependent decrease of platelet counts as well as fibrinogen @
RECOMBINANT HUMAN FACTOR VHa (rFVIIa) IN A RABBIT STASIS MODEL
Viggo Diness, Claus Bregengaard, Elisabeth Erhardtsen and Ulla Hedner Novo Nordisk A/S, Novo Alle, DK-2880 Bagsvaerd, Denmark
(Received 105.1992; accepted in revised form 22.6.1992 by Editor H.A. Vinazzer)
ABSTRACT The thrombogenicity of recombinant human FVIIa (rFVIIa) and FEIBA was studied in a rabbit stasis model. Only minor thrombus formation in isolated vein segments during 10 min. of stasis was seen following the administration of rFVIIa 100-1000 pg/kg or FEIBA 50-100 U/kg whereas both compounds caused clear thrombus formation during 30 min. of stasis. RFVIIa caused no change in platelet counts or plasma fibrinogen 3 hours after administration, whereas a small decrease of AP’IT was seen due to the direct effect of rFVIIa in this assay. In contrast, FEIBA caused a significant and dosedependent decrease of platelet counts as well as of fibrinogen, and an increase of APTT which demonstrate a general consumption of coagulation factors. In conclusion the study demonstrated a local pharmacological effect of rFVIIa in the absence of a general activation of the coagulation cascade.
INTRODUCTION It has been demonstrated that recombinant human factor VIIa (rFVIIa) is effective in the treatment of bleeding episodes in haemophilia patients with inhibitors against FVIII or FIX (1,2). Such patients are currently treated with plasma preparations containing activated coagulation factors claimed to by-pass the factor VIII/factor IX-dependent pathway of the coagulation system. This treatment is associated with the well-known thrombogenic risk of such preparations which may lead to disseminated intravascular coagulation (DIC) (3,4).
Key Words: FVIIa, DIC, rabbit stasis model
233
234
RECOMBINANT Vlla
Vol. 67, No. 2
In the present study the potential thrombogenicity of rFVIIa was examined in a rabbit stasis model. These results were compared to results obtained on the plasma-derived preparation FEIBA. Also potential interaction between rFVIIa and FEIBA was examined.
MATERIALS AND METHODS Operation: New Zealand white rabbits, females 2.5-3.3 kg body weight, were used. Anaesthesia was induced by intravenous administration of about 2 ml of pentobarbital sodium 5% via the marginal ear vein and was maintained throughout the experiment by additional administrations of about 0.1 ml per 15 min. according to effect. The rabbit was placed on the back. A PE-90 catheter was placed in a. carotis and kept open by infusion of saline. This catheter was also connected to a pressure transducer (Gould P 23 XL) and recorder (Kipp & Zonen) for continuous registration of blood pressure. A PE-50 catheter was placed in v. femoralis for administration of test compounds. Finally, v. facialis in both sides were isolated and non-occlusive ligatures were placed around the veins. Materials: Recombinant human FVIIa was produced by Novo Nordisk as previously described (5). In the final preparation, rFVIIa is the activated two-chain molecule with a specific activity of > 20,000 U/mg in a one-stage clotting assay (6). The amount of rFVIIa (mg) was determined using a sandwich ELISA based on two monoclonal antibodies (7). The freeze-dried preparation was reconstituted in sterile water to a concentration of 0.6 mg/ml which was used without further dilution. FEIBA (Immuno) (500 units), commercially available, was reconstituted in 20 ml of sterile water according to the recommendations of the manufacturer. These 20 ml were stored at -80°C until use. Immediately before intravenous administration this solution (25 U/ml) was thawed and further diluted in saline + 1% bovine serum albumin to final concentrations of about 15-18 U/ml according to the weight of the rabbit. Administration of test compounds. The experimental groups are shown in Tables l-2. A bolus of saline 1 ml/kg was given as control (group 1). RFVIIa 100, 300, and 1000 pglkg was given as a bolus during 1 min. (groups 2-4). FEIBA 50 and 1Oc’U/kg (groups 5-6) was given as an infusion over a period of 15 min. Group 7 receive? FEIBA 50 U/kg. Five minutes after the termination of FEIBA infusion rFVIIa 100 pglkg was given as a bolus. After administration of each compound the catheter was flushed with saline. The study was carried out randomized and according to Good Laboratory Practice (GLP). Stasis. Ten minutes after administration of the test compound(s) an atraumatic clamp was placed on one facial vein close to the jugular vein, thereby causing total stasis. Blood was allowed to fill the facial vein. Then a second clamp was placed on the vein, thus isolating a segment of about 1.5 cm between the two clamps. One minute later this procedure was repeated on the facial vein at the other side. After 10 minutes of stasis, the loose ligatures at one side were tightened and the vein segment was removed and inspected for the presence of thrombus formation. This procedure was repeated at the other side after 30 minutes
Vol. 67, No. 2
RECOMBINANT Vlla
235
of stasis. If present, thrombi were transferred to pre-weighed vials for determination of wet weight. Blood samples. Blood samples were taken via the catheter in a. carotis administration of the test compound(s) and 5 min. and 3 hours after, respectively.
before
Analyses. Whole blood platelet counts, plasma concentration of fibrinogen and activated partial thromboplastin time (APTT) were determined in all pretreatment and 3-hoursamples. Determination of platelet counts in EDTA-stabilized blood was carried out in an Analysis 148 C hemocounter. Plasma was prepared from blood anticoagulated with sodium citrate (final concentration 0.013 M) and stored at -80°C until analysis. Fibrinogen was determined according to the method described by I.M. Nilsson and B. Olow (8). This method includes coagulation of plasma by thrombin (Topostasin, Roche) and isolation of the fibrin coagulum. The detection limit of fibrinogen was found to be 35% of rabbit reference plasma levels. Activated partial thromboulastin time (AP’IT) was measured in a Behnk Coagulator using CepQotest (Nycomed, Oslo) according to the recommendations of the manufacturer. It was controlled whether rFVIIa or FEIBA interfered directly with the assays. RFVIIa antigen was determined in the 5-min. and 3-hour-samples from groups 2, 3 and 4. Furthermore, 10 ~1 or 30 ~1 of the injected preparation was added to 1 ml of predose plasma which was analyzed in parallel with the postdose samples. All plasma samples for determination of rFVIIa antigen were diluted 1: 10 in the ELISA dilution buffer before storage at -80°C. The ELISA assay (NovoClone, Novo Nordisk) is based on two monoclonal antibodies and does not detect endogenous FVII in the rabbit (7). The dilution buffer is a phosphate buffer, pH 7.2, which contains Tween 20 and bovine serum albumin, 1%. Calculations. Platelets (X 109/1), fibrinogen (g/l), and APTT (sec.) 3 hours after administration of test compounds were all calculated as per cent (%) of the corresponding predose value in each experiment. These relative values were used as the final results and subjected to statistical analysis. Statistics. Mann Whitney’s test was used for all statistical evaluations,
RESULTS Local thromhus formation: Results are given in Table 1. No thrombus formation during 10 or 30 min. of stasis was seen in the saline control group (Table 1). During 10 min. of stasis very little thrombus formation was seen after administration of, rFVIIa lOO1000 pg/kg, a small thrombus being formed in only 3 out of 18 vein segments. The mean thrombus weight was less than 1 mg (total range O-3.6 mg). Also FEIBA 50-100 U/kg induced little thrombus formation during 10 min. of stasis. During prolonged stasis (30 min.) rFVIIa as well as FEIBA caused clear thrombus formation in almost all vein segments. When rFVIIa 100 pglkg was given in conjunction with FEIBA 50 U/kg thrombus formation was more rapid since thrombi of approximately 10 mg were seen already after 10 min. of stasis (p < 0.01 for group 7 versus groups 2, 5 and 6).
236
RECOMBINANT Vlla
Vol. 67, No. 2
TABLE 1 Local thrombus formation in isolated vein segments. Number of thrombi and mean thrombus weight (mg) after 10 and 30 min. of stasis (n = 6) 10 min. stasis Test Group
I 1 haline 2
I
6
rFVIIa 100 pglkg
I FEIBA
100 U/kg
30 min. stasis
I
Number of thrombi
Mean weight (mg)
Number of thrombi
Mean weight (mg)
0
0
0
0
I
1
1
0.4
1
4
1
3.2
1
1
1
1.5
1
6
I
7.1
I
9.8
1
6
(
8.2
1
6
*n=5 Analyses: In the control group the predose values were: platelets 307f54 x 109/1, fibrinogen 2.4kO.3 g/l, APTT 27.4k3.2 set (mean *SE). The postdose values were calculated as per cent of the predose values in all groups (Table 2). In the saline control group platelet counts, fibrinogen, and AP’IT changed little during the experiment, the postdose values being about 90% of the predose values. Compared to the control group rFVIIa 100-1000 pg/kg caused no change of platelet counts or fibrinogen, whereas APTT was slightly shortened. In contrast, FEIBA caused a significant and dosedependent decrease of platelet counts as well as fibrinogen @
