Experimental haemorrhagic effect of two-domain non-glycosylated tissue factor pathway inhibitor compared to low molecular weight heparin.

The glycosylated multivalent three-domain Kunitz inhibitor TFPI is a natural inhibitor of tissue factor-FVIIa complex in the presence of FXa. TFPI has an experimental antithrombotic capacity indistinguishable from LMWH in a prophylactic dose, regardless of glycosylation and of the third domain. An inherited equilibrium between antithrombosis and haemorrhage exists. The aim of the study was to evaluate whether a two-domain non-glycosylated TFPI (117QTFPI1-161) has a bleeding potential in a rat gastric mucosa model. Groups; placebo, LMWH (tinzaparin) 60 and 250 anti-Xa IU/kg and 117 QTFPI1-161 1.0 and 10.0 mg/kg, given i.v. (bolus injection), randomised double dummy design. All actively treated groups significantly prolonged both the bleeding volume (493-984 microliters) and the bleeding time (10-20 min) compared to placebo (41 microliters, 2 min). It was not possible to distinguish a difference between the lower dose of LMWH and 117QTFPI1-161 in either parameter (p = 0.23-0.71). The two doses of 117QTFPI1-161 caused elevation of plasma-TFPI, 18 and 150 times baseline value. Both LMWH doses (0.6-3.2 anti-Xa IU/ml) and both 117QTFPI1-161 doses (0.2-2.7 anti-Xa IU/ml), caused significant effect in the anti-Xa assay, however 117QTFPI1-161 significantly less. Only the largest dose of 117QTFPI1-161 caused significant prolongation in the APTT assay (34 s). Both doses of LMWH caused significant prolongation (60-300 s). LMWH was the only substance to prolong the dilute-PT assay. Non-glycosylated two-domain 1.0 mg/kg TFPI, yielding supraphysiological plasma concentration, has an experimental haemorrhagic potential indistinguishable from LMWH in a prophylactic dose. The effect mediated by this type of TFPI could primarily be due to an inhibition of FXa.

Prostacyclin and thromboxane release from the vessel wall--comparison between an incubation and a perfusion model.

Prostacyclin (measured as its stable degradation product 6-keto-PGF1 alpha) and thromboxane (measured as its stable degradation product TxB2) produced by the vascular wall were measured by radioimmunoassay (RIA). Four pieces from the rabbit aorta and four from the caval vein were used. One piece was incubated in Hank's balanced salt solution (HBSS), one piece with additional indomethacin, and the other pieces were mounted in a perfusion system so that only the endothelium was exposed to the buffer solution with or without indomethacin. There was a higher release in the piece incubated in the buffer than in the piece which was perfused, indicating that not only the endothelium releases prostacyclin and thromboxane from the vascular wall. The 6-keto-PGF1 alpha/TxB2 ratio was higher in the perfused than in the incubated samples suggesting that 6-keto-PGF1 alpha release is higher in the endothelium than in the other wall layers and/or that TxB2 production is higher in the outer layers than in the inner layers. No correlation was found between the release from the incubated vessel and from the perfused vessel. There was a higher release of 6-keto-PGF1 alpha from aortas than from caval veins, when incubated or perfused, whereas there was a tendency to a higher release of TxB2 from veins than aortas.

The influence of non-ionic contrast media on the endothelium of small arteries. A comparison of metrizamide, iohexol and iopentol.

In 36 rabbits, small catheters were introduced into the central artery of one ear and into the saphenous arteries of both legs. Saline, metrizamide 170 mg I/ml, iohexol 137 mg I/ml and iopentol 137 mg I/ml, all solutions isotonic with blood, were tested. Perfusion with the test-solutions was performed with 2 ml x 3 at intervals of 5 min. In the case of iopentol, a dose-response study was performed, 1 ml x 3 and 4 ml x 3 being also tested. The solutions were administered under gentle pressure and at room temperature. Blood reflow was always observed between injections. All perfusates caused endothelial damage. Minor trauma led to endothelial cell contraction. More severe trauma increased the degree and numbers of contracted endothelial cells, frequently resulting in patches of denudation, sometimes the location for thrombus formation. In the control group only few effects on the endothelium were noted. Metrizamide caused more intimal damage at all times studied than either iohexol or iopentol. The immediate effects of saline and iopentol were quite identical, but at 2 h and 24 h iopentol caused much less intimal damage than saline.

Dextran and release of prostacyclin from ex vivo perfused rabbit arteries and veins.

The effects of dextran on the haemostatic system lead to prolongation of the bleeding time. To observe if the coating of endothelial cells by dextran influences the release of prostacyclin from the vessel walls in response to ex vivo perfusion, dextran 70 (1 g/kg b.w.) was given i.v. 15 min or 4 hours before excision of caval vein and aorta in rabbits, or was added to the perfusate. Vascular segments were perfused for 5 x 15 min, with the perfusate changed after each period and with arachidonic acid added on the last occasion. Prostacyclin was measured as its stable degradation product 6-keto-PGF1 alpha by radioimmunoassay. Arteries released more prostacyclin than did veins (p less than 0.01). Dextran given in vivo or ex vivo had no influence on release of prostacyclin from the walls of perfused arteries or veins. The anti-haemostatic properties of dextran 70, 1 g/kg b.w., thus do not seem to be mediated by prostacyclin release from vessel walls.

The haemorrhagic effect of low molecular weight heparins, dermatan sulphate and hirudin.

In a randomized, blind study the primary effect on haemostasis after intravenous administration of dermatan sulphate (DS), recombinant hirudin (r-hirudin) and four commercial low molecular weight heparins (LMWHs) (nadroparine, enoxaparin, dalteparin and tinzaparin) was investigated in rats and compared with saline (control). The tail bleeding time, the bleeding from the gastric mucosa [the mucosal bleeding time (min) and the mucosal bleeding (microliter)] as well as changes in activated partial thromboplastin time, antifactor IIa and Xa activities were investigated. DS and r-hirudin were investigated in a dose potentially suitable in thomboprophylaxis and the LMWHs in doses recommended by the manufacturers for thromboprophylaxis, adjusted to body weight. All substances significantly prolonged the mucosal bleeding time. Dalteparin, tinzaparin, DS and r-hirudin increased the mucosal bleeding when compared with controls whereas nadroparine and enoxaparin did not. The effect of r-hirudin was also significantly more pronounced compared with other treatments. Moreover, r-hirudin prolonged the tail bleeding time significantly whereas the other substances did not. The antifactor Xa activity in plasma correlated well with the given dose of the LMWHs (rs = 0.7). However, the monitored bleeding parameters in the LMWH groups did not correlate with the plasma activities of antifactor IIa or Xa. The results indicate that the tested LMWHs are not equipotent in their effect on haemostasis in this model and that antifactor IIa or Xa activities do not directly correlate with their effect on haemostasis although increased haemorrhage was observed in the LMWHs with lower antifactor Xa/antifactor IIa ratios.(ABSTRACT TRUNCATED AT 250 WORDS)

Effect of dextran 70 and saline on thrombus formation following arteriotomy and intimectomy in small arteries.

Twenty-four arteries of rabbit ears, divided into three groups of eight, were prepared and 32P-platelets injected. Arteriotomy (7 mm) and intimectomy (5 mm) were performed and in vivo platelet accumulation followed for 2 hours. Group A comprised untreated control animals, group B was treated with 17 ml saline/kg bw, and group C with 1 g dextran and 17 ml saline/kg bw (Macrodex). Significant differences in platelet accumulation were observed only between the control and Macrodex groups at 105 and 120 minutes. In the control and saline groups four of eight vessels showed poor or no patency. All Macrodex vessels showed good patency. Control and saline vessels had large amounts of red thrombotic material, except for three saline cases with small amounts. After Macrodex treatment five of eight vessels were apparently clean, while the other three showed only small amounts of red thrombotic material. Dextran seems not to influence platelet function but rather to inhibit fibrin stabilization and probably increases fibrinolysis. Vascular patency was only endangered by the formation of solid fibrin-containing red thrombi.

The influence of dextran and saline solution upon platelet behavior after microarterial anastomosis.

Microarterial anastomoses were performed on the central arteries of the ears in rabbits. An untreated control group was compared with two groups treated with saline solution and dextran 70 in saline solution respectively. Anastomotic bleeding times were prolonged in both treated groups. In vivo accumulation of 32phosphorous labelled platelets infused prior to any treatment was measured at the anastomotic sites for two hours after anastomosis. Treatment with saline solution resulted in prolonged periods of platelet accumulation--desaggregation processes and abnormally high levels in 50 per cent of the rabbits. Treatment with dextran resulted in increased quantitative accumulation of platelets at all times compared with the control groups. In some rabbits treated with dextran, platelet accumulation patterns indicated microembolization due to increased platelet thrombus fragility. Despite increased platelet accumulation in the dextran groups all vessels showed good patency and only small amounts of thrombosis material compared with the control and saline solution groups. Poor patency was registered only in one instance in the control group and two instances in the saline solution group. Using ex vivo platelet aggregometry, saline solution slightly decreased and dextran slightly increased platelet aggregability. The conclusion is that dextran does not act as an antiplatelet agent but minimizes formation of dangerous solid fibrin platelet thrombi due to a defective fragile fibrin structure adn perhaps increased fibrinolysis.

The appearance of endothelium in small arteries after treatment with 5-fluorouracil. An electron microscopic study of late effects in rabbits.

Cardiotoxicity is an unexplained toxic manifestation of 5-fluorouracil (5-FU). Its possible mechanism could be a direct cytotoxic effect on the vascular endothelium. We have tested this hypothesis in an experimental study in rabbits, using scanning and transmission electron microscopic evaluation of endothelium in small arteries (the central artery of the ear). The perfusion fixation method at physiological pressure and temperature was used. Both local and systemic effects of 5-FU on endothelium were studied 1, 3, 7, 14 and 30 days after in vivo treatment with 5-FU. Fifteen rabbits were used and five additional animals served as controls. The following parameters were evaluated: vessel wall and endothelial cell contraction, cell oedema, cytolysis, occurrence of denuded areas, platelet adhesion/aggregation and fibrin formation. For the description of each parameter, a scale of negative points (0.0-3.0) was used. We found severe cell damage with accompanying thrombus formation. The findings support the hypothesis that the thrombogenic effect of 5-FU, secondary to its direct cytotoxic effect on endothelium, is the pathophysiological mechanism behind 5-FU cardiotoxicity.