Effect of intravenous heparin infusion on thrombin-antithrombin complex and fibrinopeptide A in unstable angina.

BACKGROUND: In unstable angina, the clinical efficacy of heparin is limited in time, and recurrence of adverse events has been reported after discontinuation of the anticoagulant. METHODS: In 21 episodes of unstable angina, we used the plasma level of fibrinopeptide A (FPA) and of thrombin-antithrombin complex (TAT) to evaluate the pattern of thrombin inhibition by heparin and the effect of stopping heparin and initiating aspirin. RESULTS: At admission, the plasma level of FPA was increased: median value 3.7 ng/mL compared with 5.5 ng/mL in a control group of 20 patients with early myocardial infarction (not significant). The following findings were observed during a 4-day course of intravenous heparin infusion: (1) FPA decreased significantly 6 hours after the start of therapy; (2) FPA was lower when activated partial thromboplastic time (aPTT) was >1.5 times baseline; (3) there was a significant negative correlation between aPTT and FPA. Twenty-four hours after heparin was discontinued and aspirin initiated, a significant increase in TAT and FPA in plasma was observed. CONCLUSIONS: The results confirm ongoing fibrin formation in the active phase of unstable angina, indicate incomplete and variable inhibition of thrombin by heparin during continuous infusion, and suggest a risk of re-emergence of thrombosis (in spite of initiating aspirin) 24 hours after withdrawal of heparin. Data demonstrate a better control of thrombin activity when heparin is infused at rates that maintain aPTT at >1.5 times baseline, as currently recommended in unstable angina.

Cyclosporine-induced detachment of vascular endothelial cells initiates the intrinsic coagulation system in plasma and whole blood.

Cyclosporine A (CsA) is supposed to alter the metabolism of vascular endothelial cells, leading to a prothrombotic state. We examined by which mechanism human umbilical vein endothelial cells (HUVECs) treated with CsA would promote coagulation in human plasma and in whole blood. Treatment of HUVECs with CsA at concentrations clinically used led to dose-dependent cell detachment, with the subsequent exposure of the highly procoagulant connective tissue. As determined by scanning electron microscopy, cell counting of detached and adherent cells, and antigenic measurement of collagen exposure, HUVECs treated with 0.4 micrograms/ml CsA (or more) for 4 days exhibited significant amounts of subendothelial areas. On CsA-treated HUVEC monolayers, the clotting time of recalcified citrated platelet-rich plasma (PRP), but not platelet-poor plasma (PPP), was dose-dependently shortened. Likewise, the onset of thrombin generation was significantly earlier. Except at a high concentration of 8.0 micrograms/ml CsA, there was no procoagulant effect when PPP was used. To investigate CsA-treated HUVECs in whole blood, cells were cultivated on globular microcarriers and were incubated with nonanticoagulated whole blood. When untreated cells were used, generation of factor Xa, thrombin, and kallikrein was completely suppressed for 30 minutes. HUVEC beads treated with 0.4 and 0.8 micrograms/ml CsA, however, led to a dose-dependent generation of all three coagulation factors, with peak values at 2.5 to 5 minutes. Extrinsic activation was excluded, since CsA treatment did not induce tissue factor activity in HUVECs. Furthermore, the thrombomodulin activity of HUVECs w as not altered by CsA. In conclusion, treatment of HUVECs with CsA for 4 days at concentrations clinically used leads to the exposure of subendothelial areas that induce activation of the intrinsic coagulation in recalcified PRP and nonanticoagulated whole blood.

Rebound after cessation of oral anticoagulant therapy: the biochemical evidence.

The existence of a phenomenon of rebound hypercoagulability after cessation of oral anticoagulant therapy is controversial. The sensitive procoagulant markers for in vivo thrombin and fibrin formation are potential tools for the reassessment of the presence of each a phenomenon. We examined 19 patients anticoagulated for 6 +/- 2 months (SD, range 3-12) because of venous thromboembolism or myocardial infarction as follows: twice during stable, oral anticoagulation (INR 3.1-3.7) and then on days 1, 2, 3, 4, 5, 7, 9, 11, 13, 15, and > 30 after cessation of oral anticoagulation. Thrombin-antithrombin III complexes (TAT) and fibrinopeptide A (FPA) were measured in addition to the prothrombin times and factors II, V, VII, and X. None of the 19 patients developed clinically manifest thromboembolism within the following 9-18 months. However, the patients' TAT levels increased transiently: rising from 1.5 +/- 0.1 ng/ml (SEM) to 3.0 +/- 0.2 ng/ml on day 4 (P < 0.001), and returned to 1.7 +/- 0.1 ng/ml after day 30 (normals 1.8 +/- 0.33). 17/19 patients showed TAT peak levels above the upper limit of normal between days 3 and 11 (average: day 4), which normalized again after 30 d. 8/19 patients also had transient FPA levels above the upper normal limits ( < 1.81). We conclude that our patients increased their thrombin and fibrin formation transiently and that a subpopulation reached values consistent with a prethrombotic state.

The suppression of the coagulation of nonanticoagulated whole blood in vitro by human umbilical endothelial cells cultivated on microcarriers is not dependent on protein C activation.

Human umbilical vein endothelial cells (HUVEC) were cultivated on globular microcarriers in order to improve the endothelial cell surface to blood-volume ratio over the conventional flat bed cultures. HUVEC-beads were tested for their modulation of blood coagulation using a combination of two steps: HUVEC-beads were added into the syringe used for the venipuncture, in order to achieve immediate contact between cells and blood, and no anticoagulant was used during the incubation time of HUVEC-beads with whole blood. The coagulation initiation produced by venipuncture was almost completely suppressed as judged by thrombin measurements over a period of 60 min. The activated partial thromboplastin time showed a prolongation by a factor > 3. Direct measurements of activated protein C (APC) were negative. Moreover, inhibition of APC-generation with a monoclonal anti-human protein C antibody did not affect the anticoagulant properties of endothelial cells. Therefore the anticoagulant properties exerted by HUVEC-beads are not dependent on APC.

Evaluation of back pain secondary to spinal epidural hematoma associated with aspirin intake and a partial platelet glycoprotein Ia/IIa deficiency.

We report a case of spontaneous spinal epidural hemorrhage with three unusual features: (1) the hemorrhage was associated with aspirin ingestion and a reduced level of platelet glycoprotein Ia/IIa; (2) the patient presented with typical severe back pain but without neurologic dysfunction; and (3) the patient initially recovered without surgical decompression but suffered from recurrent epidural hematoma.

Fibrin formation and degradation in patients with arteriosclerotic disease.

BACKGROUND: The blood coagulation cascade was reported to be activated in patients with arteriosclerotic disease of the lower limbs (peripheral arterial disease, PAD). There is more thrombin and fibrin formation compared with healthy control subjects. In many studies, however, the presence of arteriosclerotic disease had not been thoroughly ruled out in the control group. Therefore, markers of the activation of the blood coagulation cascade were measured in patients with PAD and in a carefully defined control group, both groups being subjected to an exercise test. METHODS AND RESULTS: Twenty-two patients with angiographically documented PAD of grade II (Fontaine classification) and 13 control subjects in whom the presence of arteriosclerotic lesions was ruled out by noninvasive means in the carotid arteries, abdominal aorta, leg arteries, and coronary arteries took part in the study. Before and immediately after a treadmill stress test, the concentrations of prothrombin fragment F1 + 2 (F1 + 2), thrombin-antithrombin III complexes (TAT), fibrinopeptide A (FPA; this peptide was measured in spot urine also), and D-dimers were measured. Before exercise, the concentrations of F1 + 2 (1.0 +/- 0.6 versus 0.7 +/- 0.3 nmol/L), TAT (2.9 +/- 2.1 versus 1.9 +/- 0.8 micrograms/L), and D-dimers (318.2 +/- 270.1 versus 150.0 +/- 91.4 micrograms/L) were significantly higher in the patients with PAD compared with the healthy control subjects. FPA concentrations in plasma (1.9 +/- 1.0 versus 1.4 +/- 0.6 micrograms/L) and spot urine were not different, however. F1 + 2, FPA, and D-dimer concentrations correlated with the severity of the PAD as assessed by the ankle systolic blood pressure index (ABPI). The symptom-limited stress test did not lead to further activation of the blood coagulation cascade. However, concentrations of F1 + 2 (P < .001) and TAT (P < .01) after exercise correlated with the presence of ischemic changes in the stress-test ECG. CONCLUSIONS: There is evidence of enhanced thrombin formation in patients with PAD compared with an age- and sex-matched control group without clinical and sonographic evidence of arteriosclerosis. The thrombin formed, however, appears to be almost completely neutralized by antithrombin III. No direct evidence of fibrin formation was obtained, since the FPA concentrations were not different. In the patients with PAD, the higher concentrations of D-dimers are indicative of in vivo fibrinolysis. Thus, some fibrin formation must be postulated to occur in patients with arteriosclerosis.

Human endothelial cells suppress prothrombin activation in nonanticoagulated whole blood in vitro.

Human endothelial cells cultivated on polystyrene microcarrier beads were used to study endothelial anticoagulant activity in vitro. Spontaneous whole blood coagulation was inhibited by endothelial cells on microcarriers at a surface to volume ratio of 16 cm2/ml blood. Thrombin activity generated in nonanticoagulated whole blood during 1 hour and assessed by its fibrinogen clotting effect was reduced by 87% in the presence of endothelial cells. Consistent with this observation, prothrombin fragment1+2, fibrinopeptide A, and thrombin-antithrombin III-complex release during the same period of time were inhibited by 81%, 47%, and 88%, respectively. Immunoblotting analysis of cell-free supernatants derived from the same samples demonstrated that prothrombin activation was strongly suppressed in the presence of endothelial cells. Furthermore, the incubation of nonanticoagulated whole blood with endothelialized beads for only 5 minutes after venipuncture was sufficient to prevent subsequent prothrombin activation in the cell-free supernatants of the same whole blood sample after centrifugation. These findings suggest that interruption of the coagulation cascade is probably one major mechanism of endothelial anticoagulant activity in vivo.

Whole blood clot lysis: enhanced by exposure to autologous but not to homologous plasma.

A new phenomenon is described: Whole blood clots lyse faster in the plasma of the same donor than in another donor's plasma. We have confirmed this finding in 68 healthy volunteers by a standardized, pair-wise analysis and have found a mean difference in clot weights of 8.8 +/- 0.99% (SEM, p < 0.0001) after 6 h of urokinase-induced (200 U/ml) clot lysis. No difference was found in a group of 7 pairs of identical twins. Further analysis revealed that increasing concentrations of platelets in the plasma reduced the difference significantly but did not abolish it. A 1:1 mixture of autologous with homologous plasma reduced the autologous advantage by almost 50%, thus making an inhibitor unlikely. The absence of cellular components in clots of platelet-poor plasma resulted in the loss of the advantage after 2 h of lysis, but not in the early phase. We conclude that there is a clear advantage of autologous over homologous clot lysis. Potential mechanisms are discussed and include an increased affinity of enzymes for their substrates in a given individual.

Erythrocyte deformability has no influence on the rate of erythrophagocytosis in vitro by autologous human monocytes/macrophages.

Erythrocytes with decreased deformability are known to be rapidly removed from the circulation by splenic macrophages. The exact mechanism is, however, not well understood. We have analysed the phagocytosis of less-deformable erythrocytes by macrophages in vitro. Human monocytes/macrophages were isolated from peripheral blood and cultured for a total time of 6 h at 37 degrees C with 5% CO2. Autologous erythrocytes of the rhesus positive donor were rigidified by heat treatment (47 degrees C for 1 h). The change in erythrocyte deformability was assessed with a filter aspiration technique; the membrane elastic modulus was found to be increased about 2.5-fold. For controls, untreated erythrocytes and erythrocytes incubated with anti-RhD-antibodies were prepared. The rate of phagocytosis during 2 h at 37 degrees C and 5% CO2 was 0.74 +/- 0.59 (erythrocytes per monocyte/macrophage) for controls, 3.58 +/- 2.72 for anti-RhD-loaded erythrocytes and 0.82 +/- 0.74 for heat-treated erythrocytes, respectively. We conclude that decreased erythrocytes deformability does not cause an increased rate of phagocytosis by monocytes/macrophages compared to normally deformable erythrocytes in our in vitro model. This suggests that the preferential removal of rigid cells in vivo is probably not a specific process, but is due to the increased splenic transit time of rigid erythrocytes and hence longer interaction time between erythrocytes and phagocytes.

Prothrombin fragment 1 + 2, thrombin-antithrombin III-complexes and fibrinopeptide A in spontaneously clotting whole blood in vitro. Effects of heparin addition and antithrombin III deficiency.

Previous in vitro studies using spontaneously clotting whole blood revealed thrombin formation and high fibrinopeptide A (FPA) concentrations measured during incubation time. This occurred in spite of normal concentrations of thrombin antagonists present in the blood of the healthy subjects examined. However, there are several reports showing that in vivo increased thrombin-antithrombin III-complex (TAT) concentrations and relatively low FPA concentrations may occur e.g. in patients with (pre)thrombotic disorders. These in vivo findings indicate more effective thrombin inhibition by antithrombin III, with almost no fibrin formation. To find an explanation for the differences observed in vitro and in vivo, we extended the in vitro studies by measuring concentrations of prothrombin fragment 1 + 2 (F1 + 2), TAT and FPA at several time points until 30 min. Our goal was to test whether thrombin at least initially is neutralized by antithrombin III, resulting in a lack of fibrin formation, either in the absence or in the presence of heparin (0.2 and 0.5 U/ml whole blood, respectively). In the absence of heparin a simultaneous increase in the concentrations of F1 + 2, TAT and FPA was observed. Thrombin was only partially neutralized by antithrombin III and large amounts of fibrin were formed. The addition of heparin virtually suppressed thrombin formation since the F1 + 2 concentration remained low. Moreover, the small amounts of thrombin formed were neutralized by antithrombin III to a greater extent than in the absence of heparin.(ABSTRACT TRUNCATED AT 250 WORDS)

Influence of endothelial surface on flow velocity in vitro.

The influence of the endothelial surface and its interaction with various plasma components on flow velocity was investigated in vitro. The falling velocity of microcarrier beads with and without cultured porcine aortic endothelial cells in a vertical tube containing different media was recorded with a video system attached to a light microscope. The falling velocity of the beads was in agreement with values predicted from Stokes's law. Endothelium did not affect the falling velocity in various media, except in plasma, where it was slower than expected. Fibrinogen was found to have a strong surface tension effect, disturbing viscosity measurements, and to bind to endothelial cells and collagen-coated surfaces. We conclude that the endothelial surface may interact with plasma and plasma constituents such as fibrinogen in a way that affects blood flow.

[Internal bleeding in patients on oral anticoagulants]. / Die innere Blutung beim oral antikoagulierten Patienten.

Five selected case reports of patients suffering from rather unusual bleeding complications during oral anticoagulant therapy are presented. The reported frequency of bleeding during oral anticoagulation varies greatly. An unacceptably high incidence of hemorrhages has been reported in North American studies of the early 1980ies. The therapeutic target INR of 2.5-4.9 in these series is comparable to that in European studies where bleeding occurred much less frequently. We suggest that the insensitive thromboplastin reagents used in North America are unsuited to guide coumarin dosage, because too many prothrombin time values were outside the INR target range. In contrast, most prothrombin time values in European studies where a sensitive thromboplastin reagent was used, were within the target range. A recent prospective investigation by 25 Swiss practitioners showed an acceptably low bleeding complication rate (2.1 hemorrhagic complications severe enough to necessitate hospitalization per 100 patient years). Observation of contraindications, regular control of the prothrombin time using a sensitive and correctly calibrated thromboplastin, participation of practitioners and hospital laboratories at quality control exercises and consideration of drug interferences with coumarins help to reduce the incidence of hemorrhagic side effects. In case of either a PT value outside the target range or manifest bleeding, the necessary measures have to be tailored to the individual situation considering the Quick value as well as the severity and localization of hemorrhage.

Fibrinopeptide A in liver cirrhosis: evidence against a major contribution of disseminated intravascular coagulation to coagulopathy of chronic liver disease.

To test the hypothesis that disseminated intravascular coagulation contributes to hemostatic failure in liver cirrhosis, fibrinopeptide A and fibrin(ogen) degradation fragment E were measured in 69 patients with stable liver cirrhosis and compared with fibrinopeptide A and fibrin(ogen) degradation fragment E in 32 healthy subjects, 33 patients with thromboembolism, and 10 patients with hypofibrinogenemic disseminated intravascular coagulation. Mean fibrinopeptide A in cirrhosis was slightly increased compared with healthy subjects (2.4 vs. 1.8 ng/ml, p < 0.005), but fourfold lower than in thromboembolism (mean fibrinopeptide A 9.7 ng/ml; p < 0.0001), and tenfold lower than in disseminated intravascular coagulation (mean FPA 24.3 ng/ml; p < 0.0001). Single fibrinopeptide A levels in cirrhosis were within the normal range in 75% of the patients, marginally increased in 9%, and definitely increased in 16%. A definite increase in both fibrinopeptide A and fibrin(ogen) degradation fragment E, which characterized the groups of patients with thromboembolism and disseminated intravascular coagulation, was found in 10% of the cirrhotic patients. Among 17 patients with cirrhosis and hypofibrinogenemia, mean fibrinopeptide A (2.7 ng/ml) was tenfold lower compared with mean fibrinopeptide A in patients with hypofibrinogenemic disseminated intravascular coagulation (p < 0.0001), whereas the frequency of increased single fibrinopeptide A levels (29%) was not significantly different compared with the 52 cirrhotic patients without hypofibrinogenemia (single levels elevated in 23% of the cases). Moreover, the frequency of hypofibrinogenemia, thrombocytopenia, or abnormal clotting times was not significantly different in cirrhotic patients with normal fibrinopeptide A level when compared with cirrhotic patients with increased fibrinopeptide A. These findings do not support an important contribution of disseminated intravascular coagulation to coagulopathy of liver cirrhosis.

Increased thrombin-antithrombin III complexes after 1 h of physical exercise.

Concentrations of thrombin-antithrombin III (TAT) complexes in plasma were previously reported to be increased after a 100-km run, while fibrinopeptide A (FPA) concentration remained unchanged. Thus, antithrombin III appears to neutralize thrombin generated during running and prevents fibrin formation. To determine the clinical relevance of these findings, we compared the effects of exhaustive running (1 h, n = 10) on the plasma concentrations of prothrombin fragments F1 and F2, TAT, FPA, and beta-thromboglobulin with the effects of recreational jogging (1 h, n = 10) and exhaustive bicycling on an ergometer (1 h, n = 8). Prothrombin fragments F1 and F2 and TAT concentrations increased significantly in each group. The most significant increase in TAT concentration was measured in the running group (from 1.72 +/- 0.49 to 3.61 +/- 1.03 ng/ml, P < 0.001). The best correlation was found between the postexercise TAT and lactate concentrations (r = 0.62, n = 28, P < 0.001). Mean FPA concentrations after exercise did not exceed normal values in any of the three groups analyzed. An increase in beta-thromboglobulin concentration was measured in the running and in the cycling group. Thus, thrombin is formed, in particular, when associated with anaerobic metabolism, and platelets are activated during high-intensity exercise.

The influence of erythrocyte shape on suspension viscosities.

Erythrocyte shape changes are known to occur in vivo and can readily be induced in vitro. We have analysed the influence of increasing stomatocytosis produced by 0-0.64 mmol l-1 chlorpromazine and increasing echinocytosis induced with 0-120 mmol l-1 salicylate or 0-8 mmol l-1 2,4-dinitrophenol on suspension viscosities. The morphological index of each sample was determined and related to the suspension viscosity. It was found that the viscosity was increased by echinocytosis in dextran-free solutions, where no aggregation occurred. The viscosity could be normalized by retransforming echinocytes into discocytes. Under conditions with erythrocyte aggregation (suspension with 4 g dl-1 dextran 70, low shear rate: 0.1 s-1) a small degree of echinocytosis produced the highest viscosity, whereas at higher degrees of echinocytosis the ability to aggregate was reduced and the viscosity was similar to that of discocytes or stomatocytes. Erythrocytes incubated in hypotonic medium (constant cell number/volume) had a higher viscosity than cells in iso- or hypertonic medium. Severely hypotonic medium led to sphering of erythrocytes which reduced the ability of these cells to aggregate and hence decreased the viscosity of suspensions with dextran at low shear rate. The results indicate that discocytes have the lowest viscosity and thus the best oxygen transport efficiency and that iso- to hypertonicity provides a lower viscosity and better oxygen transport efficiency than hypotonicity. These results may contribute to the understanding of blood flow in health and disease.

Plasmapheresis for hyperviscosity syndrome in macroglobulinemia Waldenström and multiple myeloma: influence on blood rheology and the microcirculation.

The efficacy of plasmapheresis in improving blood flow properties in patients with hyperviscosity syndrome was studied during 22 plasmapheresis treatments in four patients with hyperviscosity syndrome (three with macroglobulinemia Waldenström, one with multiple myeloma). Immediately before and after plasmapheresis (exchange volume 3 liters) the following parameters were determined: standard hematologic parameters, serum proteins, plasma viscosity, whole blood viscosity, and blood flow velocity in finger nailfold capillaries by video microscopy. The hematocrit remained unchanged. Paraprotein concentrations were markedly reduced by plasmapheresis (average 35%). Plasma viscosity fell from 5.0 +/- 3.3 cp to 2.1 +/- 1.0 cp (p less than 0.0001, normal range 1.1 to 1.4 cp). Whole blood viscosity changed accordingly. The plasma viscosity before plasmapheresis (x) determined the drop in viscosity after plasmapheresis, according to the following regression: y = 0.97 - 0.77 x; r = 0.962, p less than 0.001. The spontaneous capillary blood cell flow velocity increased from 0.33 +/- 0.14 mm/sec to 0.55 +/- 0.21 mm/sec (p less than 0.01) and the change in spontaneous flow velocity (y) was correlated with the change in plasma viscosity (x): y = 0.02 - 0.05 x; r = 0.833, n = 7, p less than 0.05. We conclude that plasma viscosity is a major determinant of capillary blood flow and that plasmapheresis is an efficient treatment of abnormal microcirculation caused by increased plasma viscosity. Our data make it possible to predict the benefit of plasmapheresis in a given situation and contribute to a better use of this valuable method.

Effects of the oral anticoagulant phenprocoumon on blood coagulation and thrombogenesis induced by rabbit aorta subendothelium exposed to flowing human blood: role of dose and shear rate.

We investigated the effect of oral anticoagulation on thrombogenesis induced by the subendothelium of rabbit aorta. Eighteen healthy volunteers underwent a 2-week treatment with the oral coumarin preparation phenprocoumon to a target international normalized ratio (INR) of 5. By using an ex-vivo perfusion chamber system, the interaction between flowing blood and exposed subendothelium was measured at low (50 sec-1) and high (650 sec-1) wall shear rates. The low shear rate simulated blood flow in venous vessels and the high shear rate simulated blood flow in arterial vessels. Deposition of fibrin, platelets, and platelet thrombi on subendothelium was quantified by morphometric and immunologic techniques. Fibrin deposition prevailed at the low shear rate (183 +/- 57 ng/mm2 vs 46 +/- 16 ng/mm2, low vs high shear rate; mean +/- SEM; p less than 0.01). In contrast, the interaction of platelets with subendothelium was more intense at high shear rates when compared with low shear rates, as indicated by higher platelet adhesion (44% +/- 2% vs 9% +/- 1% coverage of subendothelium with platelets, p less than 0.001) and platelet thrombus volumes (3.6 +/- 0.4 microns 3/microns 2 vs 1.3 +/- 0.3 microns 3/microns 2, p less than 0.001). Fibrin deposition on subendothelium was substantially reduced even at a low intensity of anticoagulation (reduction by 60% at INR 2.1 and by 75% at INR 3.7) and was abolished after high coumarin doses (INR 5.2). In contrast, a significant inhibition of platelet thrombus formation could be achieved only by high doses of phenprocoumon (INR 5.2). Our data indicate that relatively low doses of oral anticoagulants (INR between 2 and 3.5) substantially inhibit the fibrin formation on subendothelium prevailing at venous shear conditions, whereas a high intensity of anticoagulation (INR 4 to 5) is necessary to inhibit the formation of platelet-rich thrombi prevailing at arterial shear conditions.

Effects of low and high dose oral contraceptives on blood coagulation and thrombogenesis induced by vascular subendothelium exposed to flowing human blood.

We investigated the effect of oral contraceptives with low and high estrogen concentration on blood coagulation and thrombogenesis, induced by vascular subendothelium of rabbit aorta exposed to flowing human blood. Twenty healthy women intending to take oral contraceptives were studied [1] before drug ingestion (control), and subsequently during the intake of oral contraceptives with [2] low estrogen content (20 micrograms ethinyl estradiol and 150 micrograms desogestrel per day) and [3] high estrogen content (50 micrograms ethinyl estradiol and 125 micrograms desogestrel per day). All experiments were performed between day 17 and 21 of the menstrual cycle and drug effects were studied during the third tablet cycle. Deposition of fibrin, platelets and platelet thrombi on vascular subendothelium was tested at a defined blood flow and wall shear rate (10 ml/min, 650 s-1) and was quantified by morphometrical techniques. Treatment with the low and high dose contraceptive increased the plasma levels of ethinyl estradiol (728 +/- 139 and 1438 +/- 212 vs. 0 fmol/l [low and high dose vs. control], means +/- SEM, P less than 0.001) and fibrinogen (2.3 +/- 0.1 and 2.6 +/- 0.1 vs. 2.0 +/- 0.1 g/l, P less than 0.05); and decreased antithrombin III activity (95 +/- 3 and 92 +/- 3 vs. 101 +/- 3 %, P less than 0.05). Fibrin deposition on vascular subendothelium was enhanced by the high dose contraceptive only (47 +/- 4 vs. 35 +/- 4 % coverage of the subendothelial surface with fibrin, high dose vs. control, P less than 0.05). The subendothelial deposition of platelets and platelet thrombi was not changed by contraceptive treatment. These results indicate that treatment with high dose contraceptives leads to an increase of fibrin-subendothelial interactions, whereas low dose contraceptives do not significantly alter the blood-subendothelium interactions. observed in this ex vivo model of thrombogenesis.