lnt J Clin Lab Res 22:111-114, 1992
9 Springer-Verlag 1992
Evidence that factor VII levels correlate strongly with fibrinopeptide A release: evaluation by an ex vivo method G. Mariani ~, G. Iacopino ~, and R. Pasquali-Lasagni 2 1 Thrombosis Center, Hematology, Department of Human Biopathology, University of Rome "La Sapienza," Via Chieti 7, 1-00161 Rome, Italy 2 Dipartimento di Chirurgia, Istituto Regina Elena, Viale Regina Elena 291, 1-00161 Rome, Italy
Summary. The generation of thrombin was estimated by an assay for fibrinopeptide A which was developed and employed to evaluate the relationship between factor VII and thrombin release. The a m o u n t o f fibrinopeptide A released correlated strongly, when assayed in the early stages of the reaction, with factor VII coagulant activity levels in the range 5 0 - 2 , 0 0 0 units/dl. The method was then applied to study the relationship between factor VII and fibrinopeptide A in plasma from blood donors, women on the pill and patients on oral anticoagulants (with a range o f factor VII coagulant activity from 8.5 to 600 units/dl). The overall evaluation o f the relationship between factor VII and fibrinopeptide A showed a strong correlation which was higher for factor VII coagulant activity (r=0.90) than for factor VII antigen levels (r=0.817). The regression analysis which best fitted the data was the multiplicative one, indicating that thrombin formation increases faster when factor VII coagulant activity is in the upper part of the normal range or higher. In patients on oral anticoagulants, the correlation between factor VII and fibrinopeptide A was very poor. Our data fit well with the findings o f the epidemiological studies in which high levels o f factor VII coagulant activity were shown to be associated with an increased incidence o f fatal c o r o n a r y artery disease.
C A D , but not between serum cholesterol and C A D , particularly when the end point was mortality from ischemic heart disease. A n o t h e r epidemiological study [1] has confirmed these findings in a preliminary evaluation, thus reinforcing the significance of the predictive role of FVII and F B G in ischemic heart disease. In addition, a randomized, controlled clinical study [3, 4] and some crosssectional studies [2, 9] have consistently suggested that high F V I I : C levels are associated with an increased risk of thrombosis or C A D , and that this association may be of causal significance. Miller et al. [10] demonstrated a positive association between F V I I : C and serum lipid levels, suggesting that the clotting activity of F V I I can be the expression, at least partially, o f the presence of the two-chain form (c~-VIIa). In addition, Miller et al. [9] demonstrated, in an ex vivo clotting system, the relationship between F V I I : C and an increase in fibrinopeptide A (FPA) levels in plasmas activated by cold. Moreover, Meade [5] demonstrated that increased F V I I : C strongly correlate with increased FPA levels in vivo. The aim of the present study was to evaluate the kinetics of FPA release, using an ex vivo method, and the relationship between this marker of thrombin activity and F V I I over a wide range o f F V I I levels.
Key words: Factor VII - Fibrinopeptide A - Arterial thrombosis - Activated factor VII - Extrinsic pathway
Subjects and methods
Introduction In an epidemiological study, Meade et al. [6, 7] demonstrated a strong association between factor VII coagulant activity ( F V I I : C ) , fibrinogen (FBG), serum cholesterol and coronary artery disease (CAD). In this study, the multiple regression statistical analysis indicated an independent and graded association between FVII, F B G and Offprint requests to: G. Mariani
Subjects. Plasma samples from 88 blood donors (43 males, 45 females) aged 20-58 years were evaluated for their FVII :C and FVII antigen (FVII : Ag) levels and the ability to release FPA in an ex vivo system. To better evaluate the relationship between high FVII levels and FPA release, 12 plasma samples from women taking the pill, selected for their high FVII:C levels, were also evaluated. To explore the relationship between low FVII levels and FPA release, plasma samples from 33 patients on oral anticoagulants (International Normalized Ratio [INR] values ranging from 1.53 to 5.19) were also evaluated. A pooled normal plasma (PNP) (stored at -80~ obtained by pooling plasma from 20 normal subjects was used as a reference. Reagents. A plasma from a severely affected congenitally deficient
FVII patient (FVII : C < 1 unit/d0 was used for preparation of FVII
G. Mariani et al.: Factor VII and fibrinopeptide A release
112 standards. FVII:Ag was assayed by an enzyme-linked immunosorbent assay and as a negative control for kinetic studies (ELISA) (Asserachrome FVII :Ag, Stago, Asni~res sur Seine, France) [13]. Highly purified FVII zymogen was prepared using a modification of the method of Rao and Bajaj [11]. In the reaction mixtures prepared to evaluate the generation of FX activity, FPA was measured using a commercial ELISA kit (Elisa FPA, Boehringer, Milan, Italy). As an activator for the system, suitable dilutions in 0.025 M calcium chloride (CaC12) of a human tissue thromboplastin (Thromborel S, Istituto Behring, Scoppito, Italy) were used.
Kinetic studies. Kinetic studies were performed by evaluating the generation of thrombin under different experimental conditions. Briefly, to 250 ~1of normal or test plasma an equal volume &tissue thromboplastin diluted in CaCIz was added. The reaction was stopped by adding 300 I~1of a cold (4~ 6 : 1 mixture of bentonite (80 mg/ml) anticoagulant (sodium citrate, heparin, aprotinin). The mixture was kept on melting ice, shaken for 10 min and centrifuged at 3,000 rpm; the supernanant was harvested and again submitted to the same defibrinating procedure. The resulting supernatant was finally processed in the standard FPA assay.
Statistical analysis. The regression analysis was carried out either by using a multiplicative model (y = a x b) or a linear one (y = a + b x). The regression analysis and the descriptive statistics were carried out using Statgraphics version 1.2 software.
Results
fibrin were formed. Two negative controls have routinely been used, the first a FVII-deficient plasma, the second activation of the P N P with CaCI 2 only.
Evaluation o f the relationship between FVII and FPA release First, the influence of different levels of FVII on FPA release (after 15 s incubation) was evaluated by the preparation o f F V I I standards (from 50 to 2,000 units/dl) by adding k n o w n a m o u n t s of purified FVII zymogen to the FVII-deficient plasma. F a c t o r VII : C a n d F V I I : Ag both correlated well with FPA levels (r = 0.966 for FVII : C and 0.907 for F V I I : Ag).
Relationship between FVII and FPA F V I I : C , F V I I : A g and F P A levels in blood donors, w o m e n taking the contraceptive pill and patients taking oral a n t i c o a g u l a n t s are reported in Table 1. The relationship between F P A release a n d F V I I : C levels is shown in Fig. 2. W h e n all subjects were considered, FPA correlated very well with F V I I : C ( r = 0 . 9 0 , F - r a t i o = 561.7, P < 0 . 0 0 0 0 1 ) and fairly well with F V I I : A g ( r = 0 . 8 1 7 ,
Kinetic studies To avoid the influence of clot f o r m a t i o n on the evaluation of FPA release, preliminary measurements of prot h r o m b i n times were carried out with increasing dilutions of tissue t h r o m b o p l a s t i n (data n o t shown). The influence of different dilutions of tissue t h r o m b o p l a s t i n on FPA release after different i n c u b a t i o n times was then evaluated (Fig. 1). The resulting curves were of multifactorial type and no substantial differences could be detected in the early stages of F X activation between the t h r o m b o ptastin dilutions tested (1 : 750, 1 : 1,000, 1 : 1,250). U n d e r these experimental conditions, F P A release was initiated approximately 1 0 - 1 5 s after the P N P was activated. W h e n FVII-deficient plasma was activated under the same conditions, no detectable F P A was released, either during the early stages or subsequently (Fig. 1). A 1 : 1,000 t h r o m b o p l a s t i n dilution a n d an incubation time of 15 s were then chosen for the q u a n t i t a t i o n o f t h r o m b i n generation, since u n d e r these experimental conditions, significant a m o u n t s of F P A were released by the PNP, virtually no FPA was released by the FVII-deficient plasm a and because after 20 s tiny b u t significant a m o u n t s of
2,000 !
1,500 l-
LL.
0
10
20 Time (s)
30
40
Fig. 1. Kinetics of fibrinopeptide A (FPA) release by normal plasma challenged with three different thromboplastin dilutions [1 : 750 (e), 1 : 1,000 (a), 1 : 1,250 (o)]. The open squares denote the kinetics of FPA release by FVII-deficient plasma. Data represent the median value of three determinations
Table 1. Factor VII coagulant activity (FVII : C), FVII antigen level (FVII : Ag) and fibrinopeptide A (FPA) in blood donors, women on the pill and patients on oral anticoagulantsa Subjects
n
FVII : C (units/dl)
FVII : Ag (units/dl)
Ratio C/Ag
FPA (%)
Blood donors Women on the pill Patients on oral anticoagulants
88 12 33
104.9 (22.8) 237 (163.7) 25.4 (14.7)
106.2 (17.7) 128.75 (28.5) 54.5 (17.5)
0.99 (0.16) 1.7 (0.83) 0.45 (0.17)
112.19 (64.3) 397 (370.8) 6.3 (6.6)
a Geometric mean, SD in parentheses
G. Mariani et al.: Factor VII and fibrinopeptide A release
113
15
12
9
,,
nun
9
6
0
|
~
1O0
I
I
f
I
I
200
300
400
500
600
FVll :C (units/dO
Fig. 2. FPA (ng/ml x 100) released by plasma with different factor VII coagulant activities (FVII: C) F-ratio = 263, P < 0.0001) and the FVII: C/FVII : Ag ratio (r =0.83, F-ratio = 307, P < 0.0001). Very low amounts of FPA were released from the plasma of patients on oral anticoagulants compared with the PNP (Table 1); the correlation coefficient between F V I I : C and FPA was hence very low (r = 0.276, F-ratio = 2.5, P = 0.119) as was the correlation between FPA and F V I I : A g (r=0.324, F-ratio=3.6, P=0.065) and between FPA and the F V I I : C / F V I I : A g ratio (r=0.111, F-ratio=0.39, P = 0.39).
Discussion Following the publication of the epidemiological studies in which FVII was indicated as an independent risk factor for CAD, interest in FVII, in possible methods for measurement of FVII~ and in the relationship between this factor and serum lipid concentrations increased. In particular, the relationship between fats and FVII seems to be very important, since a positive, independent association between F V I I : C and both cholesterol and triglycerides has been observed by Miller et al. [10] and Scarabin et al. [12]. In addition, Miller et al. [10] demonstrated that a diet rich in fats positively correlates with predicted high F V I I : C levels. More recently, Miller et al. [8, 9] demonstrated that a diet rich in fats, leads, even in the short term, to an increase in FVII : C levels, thus suggesting the presence of an additional, dependent association between F V I I : C and serum lipids. How can increased levels of FVII lead or predispose to arterial thrombosis, since the factor, even when activated, needs the presence of tissue factor to activate FX and FIX and hence to cause thrombin formation ? Considering the biochemical features of the tissue factor/FVII interaction it seems plausible that increased levels of F V I I : C can make the whole clotting system explosive, when even tiny amounts of tissue factor become available, possibly in fissured atheromas or through the re-
lease of activated blood cells, such as monocytes. Hence, one wonders if high levels of FVII : C can induce an increase in thrombin release and how thrombin release is related to plasma levels of FVII : C. These questions appear to be of relevance since in one of the aforementioned prospective epidemiological studies [1, 2] the patients with the highest FVII : C levels were more likely to experience fatal thrombotic events. To answer these questions we have set up a simple method, whereby plasma is activated with a highly diluted source of tissue factor (to possibly mimick the physiological situation) and subsequently assayed for FPA. Before evaluating the relationship between FVII : C levels and thrombin formation (as assayed by the release of FPA) we analyzed the kinetics of thrombin formation, in order to make the test predominantly, if not exclusively, sensitive to FVII. It appeared clear to us that the FPA first released was mainly the expression of the FVII/tissue factor interaction, whereas the rapid increment in FPA release observed thereafter (Fig. 1) was related to the activation of the other factors contributing to FX activation (FX, FIX, FV and, possibly, FVIII). There is no doubt that FVII is essential for the activation of the system, as demonstrated by the lack of thrombin release when, instead of the normal plasma, a plasma deficient in FVII was used in the reaction mixture (Fig. 1). To test the method and to evaluate its sensitivity to FVII, we created a number of FVII standards by adding known amounts of purified FVII zymogen to a FVII-deficient plasma. With these FVII standards, the amount of generated FPA was shown to correlate very well with both FVII : C and FVII : Ag. When plasma from blood donors, women taking the pill and patients on oral anticoagulants were analyzed, the relationship between FPA release and F V I I : C levels was again highly significant; less significant was the correlation with FVII : Ag. This discrepancy can be ascribed to the presence of activated FVII, at least in some subjects. This hypothesis is substantiated by the presence of a high F V I I : C / F V I I : A g ratio in several donors and in the women taking the pill, subjects in whom this ratio was maximal (Table 1). Graphically (Fig. 2), the relationship between FPA and F V I I : C can be best interpreted by a second-order regression line starting from the origin of the axis and increasing steadily, when F V I I : C levels are beyond the physiological levels. The situation is quite different in patients taking oral anticoagulants. The amount of released FPA was extremely low compared with PNP, lower than one would have expected on the basis of the FVII : C levels (Table 1). This can be explained by the fact that the other factors contributing to thrombin formation (FX, FIX, FII) are also hypo- or decarboxylated, thus contributing to the depression of the whole system. In conclusion, our results demonstrate that FVII contributes to the release of thrombin in a multiplicative rather than a linear fashion. Hence, the identification of FVII as a risk factor for arterial thrombosis receives further support. In addition, the proposed method can be considered as a model suitable for the evaluation of prothrombin activation as well as for the indirect identifi-
114 c a t i o n o f a c t i v a t e d F V I I , at least until direct m e t h o d s for assaying this serine p r o t e a s e are available.
References 1. Balleisen L, Bailey L Epping P, Schultze H, Loo J van den, Epidemiological study on factor VII, factor VIII and fibrinogen in an industrial population. I. Baseline data on the relation to age, gender, body weight, smoking, alcohol, pill using and menopause. Thromb Haemost 54:475, 1985 2. Delaker K, Hijerman I, Prydz H, A novel form of factor VII in plasma from men at risk for cardiovascular disease. Br J Haematol 61: 315, 1985 3. Henriksson P, Edhag O, Orchidectomy oestrogen for prostatic cancer: cardiovascular effects. BMJ 2:74, 1986 4. Henriksson P, Blomback M, Bratt G, Edhag O, Eriksson A, Activators and inhibitors of coagulation and fibrinolysis in patients with prostatic cancer treated with oestrogen or orchidectomy. Thromb Res 44: 783, 1986 5. Meade TW, The epidemiology of haemostatic and other variables in coronary artery disease. In: Verstraete M, Vermylen L Lijnden R, Arnout J (eds) Thrombosis and haemostasis 1987. Leuven University Press, Leuven, pp 37-60, 1987 6. Meade TW, North WS, Chakrabarti RR, Stirling Y, Haines AP, Thompson SG, Haemostatic function and cardiovascular death: early results of a prospective study. Lancet I: 1050, 1980
G. Marian et al.: Factor VII and fibrinopeptide A release 7. Meade TW, Mellows S, Brozovic M, Miller SG, Chakrabarti RR, North RS, Haines AP, Stirling Y, Imeson JD, Thompson SG, Haemostatic function and ischaemic heart disease: principal results of Northwick Park Heart Study. Lancet II: 533, 1986 8. Miller G J, Walter SJ, Stirling u Thompson SG, Esnouf MP, Meade TW, Assay of factor VII activity by two techniques: evidence for increased conversion of VII to VII a in hyperlipidemia, with possible implications for ischaemic heart disease. Br J Haematol 59: 249, 1985 9. Miller G J, Seghatchian M J, Walter S, Howarth D J, Thompson SG, Esnouf MP, Meade TW, An association between the factor VII coagulant activity and thrombin activity induced by surface/cold exposure of normal human plasma. Br J Haematol 62: 379, 1986 10. Miller G J, Seghatchian M J, Walter S, Howarth D J, Thompson SG, Esnouf MP, Meade TW', Assay of factor VII activity by two techniques: evidence for increased conversion of VII to VII~ in hyperlipidemia, with possible implications for ischaemic heart disease. Br J Haematol 59:249, 1987 11. Rao LVM, Bajaj PS, Purification of human factor VII with o-(diethylaminoethyl)-Sephadex and sulphopropyl-Sephadex chromatography. Anal Biochem 136: 357, 1984 12. Scarabin PY, Bara L, Samama M, Orssad G, Further evidence that activated factor VII is related to plasma lipids. Br J Haematol 6I: 186, 1985 13. Tirindelli MC, Mariani G, Mazzucconi MG, Iacopino G, Carbonaro M, Ghirardini A, Motta M, Mandelli F, Evaluation of factor VII antigen in factor VII congenital deficiencies with a new Elisa assay. Am J Hematol 26:313, 1987
9 Springer-Verlag 1992
Evidence that factor VII levels correlate strongly with fibrinopeptide A release: evaluation by an ex vivo method G. Mariani ~, G. Iacopino ~, and R. Pasquali-Lasagni 2 1 Thrombosis Center, Hematology, Department of Human Biopathology, University of Rome "La Sapienza," Via Chieti 7, 1-00161 Rome, Italy 2 Dipartimento di Chirurgia, Istituto Regina Elena, Viale Regina Elena 291, 1-00161 Rome, Italy
Summary. The generation of thrombin was estimated by an assay for fibrinopeptide A which was developed and employed to evaluate the relationship between factor VII and thrombin release. The a m o u n t o f fibrinopeptide A released correlated strongly, when assayed in the early stages of the reaction, with factor VII coagulant activity levels in the range 5 0 - 2 , 0 0 0 units/dl. The method was then applied to study the relationship between factor VII and fibrinopeptide A in plasma from blood donors, women on the pill and patients on oral anticoagulants (with a range o f factor VII coagulant activity from 8.5 to 600 units/dl). The overall evaluation o f the relationship between factor VII and fibrinopeptide A showed a strong correlation which was higher for factor VII coagulant activity (r=0.90) than for factor VII antigen levels (r=0.817). The regression analysis which best fitted the data was the multiplicative one, indicating that thrombin formation increases faster when factor VII coagulant activity is in the upper part of the normal range or higher. In patients on oral anticoagulants, the correlation between factor VII and fibrinopeptide A was very poor. Our data fit well with the findings o f the epidemiological studies in which high levels o f factor VII coagulant activity were shown to be associated with an increased incidence o f fatal c o r o n a r y artery disease.
C A D , but not between serum cholesterol and C A D , particularly when the end point was mortality from ischemic heart disease. A n o t h e r epidemiological study [1] has confirmed these findings in a preliminary evaluation, thus reinforcing the significance of the predictive role of FVII and F B G in ischemic heart disease. In addition, a randomized, controlled clinical study [3, 4] and some crosssectional studies [2, 9] have consistently suggested that high F V I I : C levels are associated with an increased risk of thrombosis or C A D , and that this association may be of causal significance. Miller et al. [10] demonstrated a positive association between F V I I : C and serum lipid levels, suggesting that the clotting activity of F V I I can be the expression, at least partially, o f the presence of the two-chain form (c~-VIIa). In addition, Miller et al. [9] demonstrated, in an ex vivo clotting system, the relationship between F V I I : C and an increase in fibrinopeptide A (FPA) levels in plasmas activated by cold. Moreover, Meade [5] demonstrated that increased F V I I : C strongly correlate with increased FPA levels in vivo. The aim of the present study was to evaluate the kinetics of FPA release, using an ex vivo method, and the relationship between this marker of thrombin activity and F V I I over a wide range o f F V I I levels.
Key words: Factor VII - Fibrinopeptide A - Arterial thrombosis - Activated factor VII - Extrinsic pathway
Subjects and methods
Introduction In an epidemiological study, Meade et al. [6, 7] demonstrated a strong association between factor VII coagulant activity ( F V I I : C ) , fibrinogen (FBG), serum cholesterol and coronary artery disease (CAD). In this study, the multiple regression statistical analysis indicated an independent and graded association between FVII, F B G and Offprint requests to: G. Mariani
Subjects. Plasma samples from 88 blood donors (43 males, 45 females) aged 20-58 years were evaluated for their FVII :C and FVII antigen (FVII : Ag) levels and the ability to release FPA in an ex vivo system. To better evaluate the relationship between high FVII levels and FPA release, 12 plasma samples from women taking the pill, selected for their high FVII:C levels, were also evaluated. To explore the relationship between low FVII levels and FPA release, plasma samples from 33 patients on oral anticoagulants (International Normalized Ratio [INR] values ranging from 1.53 to 5.19) were also evaluated. A pooled normal plasma (PNP) (stored at -80~ obtained by pooling plasma from 20 normal subjects was used as a reference. Reagents. A plasma from a severely affected congenitally deficient
FVII patient (FVII : C < 1 unit/d0 was used for preparation of FVII
G. Mariani et al.: Factor VII and fibrinopeptide A release
112 standards. FVII:Ag was assayed by an enzyme-linked immunosorbent assay and as a negative control for kinetic studies (ELISA) (Asserachrome FVII :Ag, Stago, Asni~res sur Seine, France) [13]. Highly purified FVII zymogen was prepared using a modification of the method of Rao and Bajaj [11]. In the reaction mixtures prepared to evaluate the generation of FX activity, FPA was measured using a commercial ELISA kit (Elisa FPA, Boehringer, Milan, Italy). As an activator for the system, suitable dilutions in 0.025 M calcium chloride (CaC12) of a human tissue thromboplastin (Thromborel S, Istituto Behring, Scoppito, Italy) were used.
Kinetic studies. Kinetic studies were performed by evaluating the generation of thrombin under different experimental conditions. Briefly, to 250 ~1of normal or test plasma an equal volume &tissue thromboplastin diluted in CaCIz was added. The reaction was stopped by adding 300 I~1of a cold (4~ 6 : 1 mixture of bentonite (80 mg/ml) anticoagulant (sodium citrate, heparin, aprotinin). The mixture was kept on melting ice, shaken for 10 min and centrifuged at 3,000 rpm; the supernanant was harvested and again submitted to the same defibrinating procedure. The resulting supernatant was finally processed in the standard FPA assay.
Statistical analysis. The regression analysis was carried out either by using a multiplicative model (y = a x b) or a linear one (y = a + b x). The regression analysis and the descriptive statistics were carried out using Statgraphics version 1.2 software.
Results
fibrin were formed. Two negative controls have routinely been used, the first a FVII-deficient plasma, the second activation of the P N P with CaCI 2 only.
Evaluation o f the relationship between FVII and FPA release First, the influence of different levels of FVII on FPA release (after 15 s incubation) was evaluated by the preparation o f F V I I standards (from 50 to 2,000 units/dl) by adding k n o w n a m o u n t s of purified FVII zymogen to the FVII-deficient plasma. F a c t o r VII : C a n d F V I I : Ag both correlated well with FPA levels (r = 0.966 for FVII : C and 0.907 for F V I I : Ag).
Relationship between FVII and FPA F V I I : C , F V I I : A g and F P A levels in blood donors, w o m e n taking the contraceptive pill and patients taking oral a n t i c o a g u l a n t s are reported in Table 1. The relationship between F P A release a n d F V I I : C levels is shown in Fig. 2. W h e n all subjects were considered, FPA correlated very well with F V I I : C ( r = 0 . 9 0 , F - r a t i o = 561.7, P < 0 . 0 0 0 0 1 ) and fairly well with F V I I : A g ( r = 0 . 8 1 7 ,
Kinetic studies To avoid the influence of clot f o r m a t i o n on the evaluation of FPA release, preliminary measurements of prot h r o m b i n times were carried out with increasing dilutions of tissue t h r o m b o p l a s t i n (data n o t shown). The influence of different dilutions of tissue t h r o m b o p l a s t i n on FPA release after different i n c u b a t i o n times was then evaluated (Fig. 1). The resulting curves were of multifactorial type and no substantial differences could be detected in the early stages of F X activation between the t h r o m b o ptastin dilutions tested (1 : 750, 1 : 1,000, 1 : 1,250). U n d e r these experimental conditions, F P A release was initiated approximately 1 0 - 1 5 s after the P N P was activated. W h e n FVII-deficient plasma was activated under the same conditions, no detectable F P A was released, either during the early stages or subsequently (Fig. 1). A 1 : 1,000 t h r o m b o p l a s t i n dilution a n d an incubation time of 15 s were then chosen for the q u a n t i t a t i o n o f t h r o m b i n generation, since u n d e r these experimental conditions, significant a m o u n t s of F P A were released by the PNP, virtually no FPA was released by the FVII-deficient plasm a and because after 20 s tiny b u t significant a m o u n t s of
2,000 !
1,500 l-
LL.
0
10
20 Time (s)
30
40
Fig. 1. Kinetics of fibrinopeptide A (FPA) release by normal plasma challenged with three different thromboplastin dilutions [1 : 750 (e), 1 : 1,000 (a), 1 : 1,250 (o)]. The open squares denote the kinetics of FPA release by FVII-deficient plasma. Data represent the median value of three determinations
Table 1. Factor VII coagulant activity (FVII : C), FVII antigen level (FVII : Ag) and fibrinopeptide A (FPA) in blood donors, women on the pill and patients on oral anticoagulantsa Subjects
n
FVII : C (units/dl)
FVII : Ag (units/dl)
Ratio C/Ag
FPA (%)
Blood donors Women on the pill Patients on oral anticoagulants
88 12 33
104.9 (22.8) 237 (163.7) 25.4 (14.7)
106.2 (17.7) 128.75 (28.5) 54.5 (17.5)
0.99 (0.16) 1.7 (0.83) 0.45 (0.17)
112.19 (64.3) 397 (370.8) 6.3 (6.6)
a Geometric mean, SD in parentheses
G. Mariani et al.: Factor VII and fibrinopeptide A release
113
15
12
9
,,
nun
9
6
0
|
~
1O0
I
I
f
I
I
200
300
400
500
600
FVll :C (units/dO
Fig. 2. FPA (ng/ml x 100) released by plasma with different factor VII coagulant activities (FVII: C) F-ratio = 263, P < 0.0001) and the FVII: C/FVII : Ag ratio (r =0.83, F-ratio = 307, P < 0.0001). Very low amounts of FPA were released from the plasma of patients on oral anticoagulants compared with the PNP (Table 1); the correlation coefficient between F V I I : C and FPA was hence very low (r = 0.276, F-ratio = 2.5, P = 0.119) as was the correlation between FPA and F V I I : A g (r=0.324, F-ratio=3.6, P=0.065) and between FPA and the F V I I : C / F V I I : A g ratio (r=0.111, F-ratio=0.39, P = 0.39).
Discussion Following the publication of the epidemiological studies in which FVII was indicated as an independent risk factor for CAD, interest in FVII, in possible methods for measurement of FVII~ and in the relationship between this factor and serum lipid concentrations increased. In particular, the relationship between fats and FVII seems to be very important, since a positive, independent association between F V I I : C and both cholesterol and triglycerides has been observed by Miller et al. [10] and Scarabin et al. [12]. In addition, Miller et al. [10] demonstrated that a diet rich in fats positively correlates with predicted high F V I I : C levels. More recently, Miller et al. [8, 9] demonstrated that a diet rich in fats, leads, even in the short term, to an increase in FVII : C levels, thus suggesting the presence of an additional, dependent association between F V I I : C and serum lipids. How can increased levels of FVII lead or predispose to arterial thrombosis, since the factor, even when activated, needs the presence of tissue factor to activate FX and FIX and hence to cause thrombin formation ? Considering the biochemical features of the tissue factor/FVII interaction it seems plausible that increased levels of F V I I : C can make the whole clotting system explosive, when even tiny amounts of tissue factor become available, possibly in fissured atheromas or through the re-
lease of activated blood cells, such as monocytes. Hence, one wonders if high levels of FVII : C can induce an increase in thrombin release and how thrombin release is related to plasma levels of FVII : C. These questions appear to be of relevance since in one of the aforementioned prospective epidemiological studies [1, 2] the patients with the highest FVII : C levels were more likely to experience fatal thrombotic events. To answer these questions we have set up a simple method, whereby plasma is activated with a highly diluted source of tissue factor (to possibly mimick the physiological situation) and subsequently assayed for FPA. Before evaluating the relationship between FVII : C levels and thrombin formation (as assayed by the release of FPA) we analyzed the kinetics of thrombin formation, in order to make the test predominantly, if not exclusively, sensitive to FVII. It appeared clear to us that the FPA first released was mainly the expression of the FVII/tissue factor interaction, whereas the rapid increment in FPA release observed thereafter (Fig. 1) was related to the activation of the other factors contributing to FX activation (FX, FIX, FV and, possibly, FVIII). There is no doubt that FVII is essential for the activation of the system, as demonstrated by the lack of thrombin release when, instead of the normal plasma, a plasma deficient in FVII was used in the reaction mixture (Fig. 1). To test the method and to evaluate its sensitivity to FVII, we created a number of FVII standards by adding known amounts of purified FVII zymogen to a FVII-deficient plasma. With these FVII standards, the amount of generated FPA was shown to correlate very well with both FVII : C and FVII : Ag. When plasma from blood donors, women taking the pill and patients on oral anticoagulants were analyzed, the relationship between FPA release and F V I I : C levels was again highly significant; less significant was the correlation with FVII : Ag. This discrepancy can be ascribed to the presence of activated FVII, at least in some subjects. This hypothesis is substantiated by the presence of a high F V I I : C / F V I I : A g ratio in several donors and in the women taking the pill, subjects in whom this ratio was maximal (Table 1). Graphically (Fig. 2), the relationship between FPA and F V I I : C can be best interpreted by a second-order regression line starting from the origin of the axis and increasing steadily, when F V I I : C levels are beyond the physiological levels. The situation is quite different in patients taking oral anticoagulants. The amount of released FPA was extremely low compared with PNP, lower than one would have expected on the basis of the FVII : C levels (Table 1). This can be explained by the fact that the other factors contributing to thrombin formation (FX, FIX, FII) are also hypo- or decarboxylated, thus contributing to the depression of the whole system. In conclusion, our results demonstrate that FVII contributes to the release of thrombin in a multiplicative rather than a linear fashion. Hence, the identification of FVII as a risk factor for arterial thrombosis receives further support. In addition, the proposed method can be considered as a model suitable for the evaluation of prothrombin activation as well as for the indirect identifi-
114 c a t i o n o f a c t i v a t e d F V I I , at least until direct m e t h o d s for assaying this serine p r o t e a s e are available.
References 1. Balleisen L, Bailey L Epping P, Schultze H, Loo J van den, Epidemiological study on factor VII, factor VIII and fibrinogen in an industrial population. I. Baseline data on the relation to age, gender, body weight, smoking, alcohol, pill using and menopause. Thromb Haemost 54:475, 1985 2. Delaker K, Hijerman I, Prydz H, A novel form of factor VII in plasma from men at risk for cardiovascular disease. Br J Haematol 61: 315, 1985 3. Henriksson P, Edhag O, Orchidectomy oestrogen for prostatic cancer: cardiovascular effects. BMJ 2:74, 1986 4. Henriksson P, Blomback M, Bratt G, Edhag O, Eriksson A, Activators and inhibitors of coagulation and fibrinolysis in patients with prostatic cancer treated with oestrogen or orchidectomy. Thromb Res 44: 783, 1986 5. Meade TW, The epidemiology of haemostatic and other variables in coronary artery disease. In: Verstraete M, Vermylen L Lijnden R, Arnout J (eds) Thrombosis and haemostasis 1987. Leuven University Press, Leuven, pp 37-60, 1987 6. Meade TW, North WS, Chakrabarti RR, Stirling Y, Haines AP, Thompson SG, Haemostatic function and cardiovascular death: early results of a prospective study. Lancet I: 1050, 1980
G. Marian et al.: Factor VII and fibrinopeptide A release 7. Meade TW, Mellows S, Brozovic M, Miller SG, Chakrabarti RR, North RS, Haines AP, Stirling Y, Imeson JD, Thompson SG, Haemostatic function and ischaemic heart disease: principal results of Northwick Park Heart Study. Lancet II: 533, 1986 8. Miller G J, Walter SJ, Stirling u Thompson SG, Esnouf MP, Meade TW, Assay of factor VII activity by two techniques: evidence for increased conversion of VII to VII a in hyperlipidemia, with possible implications for ischaemic heart disease. Br J Haematol 59: 249, 1985 9. Miller G J, Seghatchian M J, Walter S, Howarth D J, Thompson SG, Esnouf MP, Meade TW, An association between the factor VII coagulant activity and thrombin activity induced by surface/cold exposure of normal human plasma. Br J Haematol 62: 379, 1986 10. Miller G J, Seghatchian M J, Walter S, Howarth D J, Thompson SG, Esnouf MP, Meade TW', Assay of factor VII activity by two techniques: evidence for increased conversion of VII to VII~ in hyperlipidemia, with possible implications for ischaemic heart disease. Br J Haematol 59:249, 1987 11. Rao LVM, Bajaj PS, Purification of human factor VII with o-(diethylaminoethyl)-Sephadex and sulphopropyl-Sephadex chromatography. Anal Biochem 136: 357, 1984 12. Scarabin PY, Bara L, Samama M, Orssad G, Further evidence that activated factor VII is related to plasma lipids. Br J Haematol 6I: 186, 1985 13. Tirindelli MC, Mariani G, Mazzucconi MG, Iacopino G, Carbonaro M, Ghirardini A, Motta M, Mandelli F, Evaluation of factor VII antigen in factor VII congenital deficiencies with a new Elisa assay. Am J Hematol 26:313, 1987
