A63 Quantifying the dynamics of evolutionary rates through time

The availability of evolutionary rate estimates in recent years led to the observation that they may depend on the time-scale on which they are measured. Specifically, RNA virus evolutionary rates are frequently estimated to be low towards the past and high towards the present. This time-dependent rate phenomenon (TDRP) has important implications for evolutionary studies as it could severely bias divergence time estimates. While recent studies are providing insights into the relationship between viral evolutionary rate and time, formal probabilistic models to draw inference under TDRP scenarios remain lacking. Here, we adopt epoch-modelling to develop a Bayesian model of discrete rate changes through time in an unknown evolutionary history and combine this with a log-linear parameterization of rates as a function of times in the past. We provide an implementation for nucleotide substitution rates as well as for nonsynonymous rates change in a codon substitution model. Using a foamy virus dataset for which internal node calibrations can be applied based on host-virus co-divergence, we estimate a significant decline in evolutionary rates as a function of time into the past for nucleotide substitutions as well as for non-synonymous substitutions in a codon model. We also estimate a deep evolutionary history for primate Lentiviruses by combining an HIV-1 group M node calibration and a biogeographic calibration for viruses in drill monkeys in the TDRP model. Our analyses lead to the conclusion that studies of evolutionary timescales require a reconsideration of substitution rates, in either codon and nucleotide substitution model, as a dynamic feature of molecular evolution.

Genomic and epidemiological monitoring of yellow fever virus transmission potential

The yellow fever virus (YFV) epidemic in Brazil is the largest in decades. The recent discovery of YFV in Brazilian Aedes species mosquitos highlights a need to monitor the risk of reestablishment of urban YFV transmission in the Americas. We use a suite of epidemiological, spatial, and genomic approaches to characterize YFV transmission. We show that the age and sex distribution of human cases is characteristic of sylvatic transmission. Analysis of YFV cases combined with genomes generated locally reveals an early phase of sylvatic YFV transmission and spatial expansion toward previously YFV-free areas, followed by a rise in viral spillover to humans in late 2016. Our results establish a framework for monitoring YFV transmission in real time that will contribute to a global strategy to eliminate future YFV epidemics.

In Vitro Inhibition of Platelet Aggregation by Bile Salts

In a series of in vitro experiments, the influence of bile salts on platelet aggregation by ADP or by collagen and on serotonin-14C release by collagen, was studied.

Sodium salts of the following bile acids showed a clear inhibitory effect: glycochenodeoxycholic acid, taurochenodeoxycholic acid, glycocholic acid, taurocholic acid and cholic acid. Lowering the pH of the platelet-rich plasma resulted in decreased platelet aggregation by ADP and by collagen. Bile salts further enhanced the inhibitory effect of pH change. In contrast the sodium salt of chenodeoxycholic acid was the sole aggregation inducing bile salt we studied. Relating the above studies to the clinical situation of upper gastroduodenal mucosal haemorrhage, we suggest that biliary reflux as well as the acidic environment may contribute to a poor haemostatic response by impaired platelet aggregation in the upper gastrointestinal tract of otherwise normal human subjects.

Granulocyte Elastase, Tumor Necrosis Factor-α and Urokinase Levels as Prognostic Markers in Severe Infection

We have examined the prognostic value of the levels in the blood of granulocyte elastase-α1-proteinase inhibitor (E-α1-PI) complex, tumor necrosis factor-α (TNF-α) and urokinase-type plasminogen activator (u-PA) in 35 patients with severe infection upon admission to an Intensive Care Unit. Fourteen patients died.

No differences for E-α1-PI complex were found between survivors and nonsurvivors, but in all patients the levels on admission were eight-fold higher than the reference value.

TNF-α levels, measured by immunoassay, on admission were four times higher in the nonsurvivors than in the survivors (p = 0.0003) and correlated with the severity of the disease (APACHE II score, r = 0.43, p <0.05). TNF-α was not detectable by bioassay.

Total u-PA antigen (u-PA Ag), plasmin-activatable single-chain u-PA (scu-PA) and inactive, nonactivatable u-PA (u-PA#) were on admission all two-fold higher in the nonsurvivors (p = 0.0006, 0.003 and 0.0003, respectively), while normal in the survivors. In both, survivors and nonsurvivors, the ratio between scu-PA and u-PA Ag was significantly decreased (p <0.001, compared to a reference group of healthy volunteers), indicative for enhanced conversion of scu-PA to active two-chain u-PA (tcu-PA) and inactive u-PA# during severe infectious disease. tcu-PA was detected in nine of the 35 patients, while virtually undetectable in controls. scu-PA correlated with the Child-Pugh score on admission (r = 0.42, p <0.05). TNF-α correlated with u-PA Ag (r = 0.38, p <0.05) and with u-PA# (r = 0.47, p <0.01).

In a stepwise logistic regression analysis, documentation of infection and plasma levels of u-PA Ag contributed most significantly to prediction of patient outcome. Serum levels of TNF-α did not. These results suggest that, in addition to a number of other clinical and laboratory parameters, u-PA Ag can be used as a prognostic marker in patients with severe infection admitted to an Intensive Care Unit.

Inhibitory Effect of Acetylsalicylic Acid on Human Platelet Function in Normal Volunteers and in Women Using a Combined Oral Contraceptive Regime

Six parameters related to the release reaction were measured simultaneously in 10 human volunteers prior to the intake of one single dose of 1 g acetylsalicylic acid and 1, 4, 5, 6 and 7 days later: ΔE5 with diluted collagen, ΔE10 with Thrombofax® and the serotonin-14C release by undiluted and diluted collagen, by Thrombofax and by bovine plasma. The duration of the inhibitory effect varied according to the test used. It was the most prolonged (through the 7th day) if serotonin-14C release by diluted collagen was measured.

A systematic investigation of the platelet release reaction in women taking a combined oral contraceptive was also performed. There were no statistically significant differences from a control group. No difference in acetylsalicylic acid sensitivity, measured 24 hours after intake of 1 g of aspirin, could be demonstrated.

Fibrinolysis and Coagulation in Patients with Infectious Disease and Sepsis

Sepsis is often associated with hemostatic dysfunction. This study aimed to relate changes in fibrinolysis and coagulation parameters to sepsis and sepsis outcome. Urokinase-type plasminogen activator (u-PA) antigen, tissue-type plasminogen activator (t-PA) antigen and activity, plasminogen activator inhibitor (PAI) type 1 antigen, PAI activity, antithrombin (AT) III activity, and protein C activity were measured in 24 patients suffering from sepsis or septic shock and the results were compared with those observed in 30 non-sepsis patients with severe infectious disease. The u-PA level was markedly increased in plasma of sepsis patients as compared to non-sepsis patients (11.5 ± 9.4 versus 1.6 ± 1.5 ng/ml, p <0.0001). PAI-1 antigen and t-PA activity showed a significant increase in sepsis patients (320 ± 390 ng/ml versus 120 ± 200 ng/ml, and 3.0 ± 3.6 IU/ml versus 1.0 ± 0.7 IU/ml, respectively, p <0.01). AT III was decreased in sepsis patients (58 ± 28% in sepsis versus 79 ± 26% in severe infectious disease, p <0.01) as was protein C (30 ± 18% versus 58 ± 27%, p <0.001). No significant difference was found for t-PA antigen nor for PAI activity. Nonsurvivors of sepsis were distinguished mainly by a high u-PA antigen level and increased t-PA activity. It is concluded that plasma u-PA antigen showed the strongest significant difference, among the parameters evaluated, between sepsis and severe infection. u-PA antigen may be of prognostic value in patients admitted to the medical intensive care unit for severe infectious disease.

Limitations of the Activated Partial Thromboplastin Time for Determination of Plasma Heparin Concentration and Half-Life in Human Subjects

In vivo heparin concentrations and half-life were estimated in 20 human subjects, after a single intravenous injection.

The method, used for heparin determination, is based on the activated partial thromboplastin time. The test plasmas are diluted in plasma of the same subject obtained before heparin administration. In this way the heparin concentrations are reduced to measurable levels between 0.1 I. U./ml and 0.8 I. U./ml. Indeed a linear relationship between log activated partial thromboplastin time and heparin concentrations from 0.1 to 0.8 I. U./ml can be computed. Different half-lives were found in 13 of the 20 subjects examined when different dilution factors were used. This seems an important limitation for the examination of the pharmacokinetics of heparin.

Identification of two different mutations causing protein S deficiency in two unrelated Belgian families using a nonisotopic scanning and sequencing method

Hereditary protein S deficiency is a risk factor for developing recurrent venous thromboembolic disease and is caused by a defect in the protein S 1 (PROS1) gene. Identification of the mutation in the PROS1 gene can overcome diagnostic uncertainty in family members with borderline protein S levels. We describe a novel nonisotopic method for molecular diagnosis of protein S deficiency, using fluorescein-labeled amplification and sequencing primers. As a first step, all exons of the PROS1 gene are selectively amplified, and heteroduplex analysis is performed. As a second step, all exons are analyzed by direct sequencing. Using this method, we have characterized the molecular defect in two Belgian families with hereditary protein S deficiency type I: a frameshift mutation in exon XIV (1881insTC) and a missense mutation caused by a T-to-C transition, resulting in substitution of Leu405 by Pro (L405P).

Intra-individual variability of fibrinogen levels

Elevated fibrinogen concentrations are recognized as playing an important role in the pathogenesis of atherosclerosis. In the framework of a risk factor survey in 342 middle-aged working men, screened twice over a period of five months, plasma fibrinogen levels were found to be fairly unstable as large discrepancies between both measurements were observed. Due to a substantial proportion of within-person variability, the reliability coefficient was only R = 0.56. Repeatability was highest in higher educated and physically more active men. Our data suggest that the impact of elevated fibrinogen levels on the development of ischemic heart disease and stroke, is likely to be under-estimated.

Comparison of a recombinant thromboplastin with Thrombotest for oral anticoagulant control

Thrombotest results expressed in international normalized ratio (INR) values, obtained in 108 patients on oral anticoagulant treatment, were compared with prothrombin time (PT) results with a recombinant thromboplastin. The former results were obtained on an Amelung coagulometer, the latter on a photo-optical instrument. Using the Thrombotest method, performed within 2 h after sampling as the reference method, a first group of 63 patients had an INR value between 2 and 4. This group was considered as adequately anticoagulated and served as a true positive population in further analysis. The remaining 45 patients (true-negative group) had an INR value below 2 or higher than 4 and could thus be considered as inadequately anticoagulated. Using these definitions, a sensitivity of 86% and a specificity of 96% could be calculated for the PT with the recombinant thromboplastin. All tests from patients on oral anticoagulant treatment were also performed after 24 h storage of the blood or plasma samples at room temperature. When we compared the reference Thrombotest results with those of the late Thrombotest and the late PT recombinant thromboplastin, sensitivities of 86 and 86% as well as specificities of 91 and 96% were found, respectively. In conclusion, PT with a recombinant thromboplastin on a photo-optical instrument, even after prolonged storage of the plasma samples at room temperature, can be considered as suitable for oral anticoagulation control.

Relation of fibrinogen to lifestyles and to cardiovascular risk factors in a working population

BACKGROUND

The association between fibrinogen and smoking behaviour, age, body mass index, blood pressure, heart rate and plasma lipid profile, was assessed in a cohort of middle-aged working men.

METHODS

Seven hundred and forty five subjects were examined as part of a health intervention programme at the work-site. Nine subjects were excluded from the study because of prevalent diabetes. Correlates of plasma fibrinogen concentrations were evaluated through univariate and multivariate methods.

RESULTS

In multiple regression analysis fibrinogen correlated with age, smoking behaviour, apolipoprotein B (apo B) apolipoprotein A-I (apo A-I) and lipoprotein (a) (Lp(a)) levels, which together explained 11% of the variation in fibrinogen concentration. From this model, fibrinogen concentration was associated with an increase of 13.6 mg/dl for every 10 years' increase in age, 28.2 mg/dl if a person smoked, and 4.6 mg/dl and 3.8 mg/dl with a 20 mg/dl rise in respectively apo B and Lp(a). A 20 mg/dl increase in apo A-I concentrations was estimated to be associated with a 6.0 mg/dl lower fibrinogen level.

CONCLUSION

The data indicate that both biochemical and lifestyle factors are related to the plasma fibrinogen concentration; these interactions may explain partly the relationship between fibrinogen and cardiovascular disease.

Acquired haemophilia in the elderly

The occurrence of an acquired inhibitor of F VIII coagulant activity is a rare cause of a sometimes important bleeding diathesis. Antibodies against F VIII:C can spontaneously occur, mostly in elderly patients. Four elderly patients with such a typical clinical and biological syndrome are reported. No identifiable underlying disease was found and they responded favourably to an immunosuppressive treatment. In one patient plasmapheresis was successful.

Coagulation system activation and increase of D-dimer levels in peripheral arterial occlusive disease

The aim of the present study was to document coagulation system activation and basal fibrinolysis in peripheral arterial occlusive disease (PAOD) at stage II of Fontaine's classification. In 34 patients, prothrombin fragment (F1 + 2), thrombin-antithrombin III complexes (TAT), and D-dimer concentrations were evaluated before and after a standard treadmill test. Basal levels in PAOD of F1 + 2 (1.25 +/- 0.19 nmol/liter) and of TAT (3.34 +/- 0.35 micrograms/liter) were significantly increased compared to those obtained in age- and sex-matched healthy controls (0.68 +/- 0.06 nmol/liter and 2.30 +/- 0.33 micrograms/liter, respectively), showing baseline activation of the clotting cascade. A secondary activation of the fibrinolytic system was evidenced by the highly significant increase of basal D-dimers (719 +/- 99 ng/dl in PAOD vs. 229 +/- 37 ng/dl in controls). Treadmill exercise failed to increase the study parameters significantly further. Walking distance (583 +/- 40 m) was correlated with the preexercise ankle to brachial systolic blood pressure ratio (r = 0.485, P < 0.005) and inversely with the level of D-dimers (r = -0.425, P < 0.02). Under baseline conditions, the latter parameter was correlated as well with the antigen concentration of urokinase-type plasminogen activator (u-PA; r = 0.503, P < 0.002). These results indicate that stage II PAOD is characterized by an activation of the clotting cascade in baseline conditions evidenced by increased F1 + 2 and TAT. A secondary activation of the fibrinolytic system with increased u-PA antigen levels accounts for the elevated D-dimers. Treadmill exercise was unable to increase these parameters further.

Granulocyte elastase, tumor necrosis factor-alpha and urokinase levels as prognostic markers in severe infection

We have examined the prognostic value of the levels in the blood of granulocyte elastase-alpha 1-proteinase inhibitor (E-alpha 1-PI) complex, tumor necrosis factor-alpha (TNF-alpha) and urokinase-type plasminogen activator (u-PA) in 35 patients with severe infection upon admission to an Intensive Care Unit. Fourteen patients died. No differences for E-alpha 1-PI complex were found between survivors and nonsurvivors, but in all patients the levels on admission were eight-fold higher than the reference value. TNF-alpha levels, measured by immunoassay, on admission were four times higher in the nonsurvivors than in the survivors (p = 0.0003) and correlated with the severity of the disease (APACHE II score, r = 0.43, p less than 0.05). TNF-alpha was not detectable by bioassay. Total u-PA antigen (u-PA Ag), plasmin-activatable single-chain u-PA (scu-PA) and inactive, nonactivatable u-PA (u-PA#) were on admission all two-fold higher in the nonsurvivors (p = 0.0006, 0.003 and 0.0003, respectively), while normal in the survivors. In both, survivors and nonsurvivors, the ratio between scu-PA and u-PA Ag was significantly decreased (p less than 0.001, compared to a reference group of healthy volunteers), indicative for enhanced conversion of scu-PA to active two-chain u-PA (tcu-PA) and inactive u-PA# during severe infectious disease. tcu-PA was detected in nine of the 35 patients, while virtually undetectable in controls. scu-PA correlated with the Child-Pugh score on admission (r = 0.42, p less than 0.05).(ABSTRACT TRUNCATED AT 250 WORDS)

Comparison of the fibrinolytic response to desmopressin acetate (DDAVP) infusion versus venous occlusion in patients with coronary artery disease

A 20-min venous occlusion and a desmopressin acetate (DDAVP, 0.4 microgram.Kg-1, 15 min) infusion test have been compared to evaluate fibrinolytic capacity in proven coronary artery disease. Basal values of plasma tissue plasminogen activator (t-PA) activity and antigen and of t-PA inhibitor (PAI) were normally distributed in this patient group. However, after both stimuli, highly significant (P less than 0.005) increases of t-PA antigen and activity have been observed. Renormalization of t-PA levels after DDAVP infusion occurred more rapidly for antigen (120 min) than for activity concentrations (greater than 240 min). A DDAVP infusion test seems more appropriate for evaluation of the fibrinolytic capacity as the induced decrease in PAI level was significant (P less than 0.005) and lasted for at least 240 min, while it was not significant for the venous occlusion test.

Fibrinolysis and coagulation in patients with infectious disease and sepsis

Sepsis is often associated with hemostatic dysfunction. This study aimed to relate changes in fibrinolysis and coagulation parameters to sepsis and sepsis outcome. Urokinase-type plasminogen activator (u-PA) antigen, tissue-type plasminogen activator (t-PA) antigen and activity, plasminogen activator inhibitor (PAI) type 1 antigen, PAI activity, antithrombin (AT) III activity, and protein C activity were measured in 24 patients suffering from sepsis or septic shock and the results were compared with those observed in 30 non-sepsis patients with severe infectious disease. The u-PA level was markedly increased in plasma of sepsis patients as compared to non-sepsis patients (11.5 +/- 9.4 versus 1.6 +/- 1.5 ng/ml, p less than 0.0001). PAI-1 antigen and t-PA activity showed a significant increase in sepsis patients (320 +/- 390 ng/ml versus 120 +/- 200 ng/ml, and 3.0 +/- 3.6 IU/ml versus 1.0 +/- 0.7 IU/ml, respectively, p less than 0.01). AT III was decreased in sepsis patients (58 +/- 28% in sepsis versus 79 +/- 26% in severe infectious disease, p less than 0.01) as was protein C (30 +/- 18% versus 58 +/- 27%, p less than 0.001). No significant difference was found for t-PA antigen nor for PAI activity. Nonsurvivors of sepsis were distinguished mainly by a high u-PA antigen level and increased t-PA activity. It is concluded that plasma u-PA antigen showed the strongest significant difference, among the parameters evaluated, between sepsis and severe infection. u-PA antigen may be of prognostic value in patients admitted to the medical intensive care unit for severe infectious disease.

Coagulation and fibrinolytic parameters in patients undergoing total hip replacement: influence of the anaesthesia technique

Eighty patients undergoing total hip replacement (THR) were randomly allocated to three groups. Group I (n = 29) received general anaesthesia, Group II (n = 29) epidural anaesthesia and Group III (n = 22) the same epidural as Group II and the same general anaesthesia as Group I but with a lower isoflurane concentration. Prothrombin time (PT), activated thromboplastin time (APTT), fibrinogen (FG), plasminogen (PG), antithrombin III (AT III), protein C (Proc C), alpha-2-antiplasmin (alpha 2AP), Factor VIII coagulating activity (F VIII:C), von Willebrand factor antigen (vWF:Ag), von Willebrand ristocetin cofactor (vWF:Rcof), tissue plasminogen activator (tPA) as antigen and activity were measured before induction (A), at the end of surgery (B), on the first postoperative morning (C) and 7 days postoperatively (D). The most relevant finding was that AT III was equally depressed immediately after surgery in all groups, but returned to normal significantly faster in the epidural group (mean values at C: 96.2% in Group I, 104.1% in Group II, 92.7% in Group III). The faster return to normal of AT III after epidural anaesthesia could be one of the mechanisms responsible for the beneficial effect of this technique on the prevention of thromboembolic complications.

Bromocriptine in an injectable retard form for puerperal lactation suppression: comparison with Estandron prolongatum

Intramuscular injection of a single 50-mg dose of long-acting bromocriptine microspheres was compared with a single intramuscular dose of an estradiol/testosterone ester combination in a single-blind, randomized study of 54 subjects. Bromocriptine was significantly more effective than the steroid drug in preventing milk flow, and rebound lactation was not observed in any bromocriptine-treated patients. Neither group showed deleterious side effects or significant biologic changes in coagulation parameters. There were no blood pressure or electrocardiographic alterations. Postpartum prolactin suppression was more intense after bromocriptine administration than after steroid therapy.

Coagulant, fibrinolytic, and aggregating activity in ascitic fluid

Ascitic fluid samples from 14 subjects with liver cirrhosis and from 13 patients with malignancy were investigated. Activated FX was present in ascitic fluid in small quantities with a mean value of 8.7 10(-3) IU/ml. The mean thrombin activity was 70.8 10(-3) IU/ml and the mean plasmin activity was 449.6 10(-3) CU/ml. High levels of fibrin/fibrinogen degradation products (mean 75.4 micrograms/ml) and of antithrombin III (mean 43.4%) were found. No statistically significant differences between values in liver cirrhosis and in malignancy were found. In 15 of 17 experiments 10-fold concentrated ascitic fluid caused irreversible platelet aggregation and [14C] serotonin release of normal platelet-rich plasma similar to collagen. The aggregating effect disappeared after addition of collagenase. These results do not support the concept that the coagulopathy after peritoneovenous shunting is a result of direct and rapid intravenous infusion of procoagulant substances. They rather point to a central role of collagen present in ascitic fluid.

Effects of oxametacin on coumarin anticoagulation and on platelet function in humans

In 12 chronically anticoagulated patients the administration of 1-p-chlorobenzoyl-5-methoxy-2-methyl-3-indoacetohydroxamic acid (oxametacin, Flogar), three times 100 mg a day, decreased the thrombotest percentage from a mean of 11.2% to a mean of 8.3% after one week and of 7.8% after two weeks of treatment. A potentiating effect of oxametacin was also found when the prothrombin time or the activated partial thromboplastin time were used as parameters. In one third of our patients an adaptation of the coumarin dose or even an interruption in the warfarin administration was necessary. Although no severe bleedings occurred, care should be taken in prescribing oxametacin to anticoagulated patients. In a second part of this study we compared the ex vivo effects of a single oral dose of 1 g of acetylsalicylic acid, 50 mg of indometacin and 100 mg of oxametacin in human volunteers. Platelet aggregation and 5HT--14C release induced by collagen. Thrombofax, ADP, adrenaline (epinephrine), bovine plasma and ristocetin were measured before, 1 and 24 h after drug administration. A clear-cut inhibitory effect as induced by acetylsalicylic acid was not found for oxametacin.

In vitro inhibition of platelet aggregation by bile salts

In a series of in vitro experiments, the influence of bile salts on platelet aggregation by ADP or by collagen and on serotonin-14C release by collagen, was studied. Sodium salts of the following bile acids showed a clear inhibitory effect: glycochenodeoxycholic acid, taurochenodeoxycholic acid, glycocholic acid, taurocholic acid and cholic acid. Lowering the pH of the platelet-rich plasma resulted in decrease platelet aggregation by ADP and by collagen. Bile salts further enhanced the inhibitory effect of pH change. In contrast the sodium salt of chenodeoxycholic acid was the sole aggregation inducing bile salt we studied. Relating the above studies to the clinical situation of upper gastroduodenal mucosal haemorrhage, we suggest that biliary reflux as well as the acidic environment may contribute to a poor haemostatic response by impaired platelet aggregation in the upper gastrointestinal tract of otherwise normal human subjects.

Perinatal hazards of chronic antenatal tocolysis with indomethacin

Indomethacin was administered from the 20th to the 34th week of gestation to 51 women (59 fetuses) in whom bed rest and a beta-mimetic compound (ritodrine) had failed to stop preterm labor. No serious maternal side effects were observed. Of the 8 perinatal deaths, 5 were not and 3 were possibly related to the prostaglandin synthesis inhibitor. The sole case of serious neonatal morbidity is not considered to have had a causal relationship with the indomethacin treatment.

Antihaemophilic factor A activity, F VIII-related antigen and von Wilebrand factor in hepatic cirrhosis

F VIII activity, F VIII-related antigen and von Willebrand factor were measured in 46 patients with hepatic cirrhosis and in 30 normal individuals. These parameters were significantly higher in hepatic cirrhosis than in the controls. Linear relationships between F VIII activity and F VIII-related antigen and between F VIII-related antigen and von Willebrand factor were found in patients with hepatic cirrhosis as well as in normal individuals. However, in both groups no relationship between F VIII activity and von Willebrand factor was present. The existence of a low-grade intravascular coagulation in hepatic cirrhosis may be postulated but more information about the metabolism of F VIII protein is needed before such a statement can be proven.

Inhibitory effect of acetylsalicylic acid on human platelet function in normal volunteers and in women using a combined oral contraceptive regime

Six parameters related to the release reaction were measured simultaneously in 10 human volunteers prior to the intake of one single dose of 1 g acetylsalicylic acid and 1, 4, 5, 6 and 7 days alter: deltaE5 with diluted collagen, deltaE10 with Thrombofax and the serotonin-14C release by undiluted and diluted collagen, by Thrombofax and by bovine plasma. The duration of the inhibitory effect varied according to the test used. It was the most prolonged (through the 7th day) if serotonin-14C release by diluted collagen was measured. A systematic investigation of the platelet release reaction in women taking a combined oral contraceptive was also performed. There were no statistically significant differences from a control group. No difference in acetylsalicylic acid sensitivity, measured 24 hours after intake of 1 g of aspirin, could be demonstrated.