A case of acquired FXIII deficiency with severe bleeding symptoms.

Acquired factor XIII (FXIII) deficiency due to an autoantibody against FXIII is a very rare, yet potentially life-threatening bleeding disorder. As the standard coagulation tests (prothrombin time and activated partial thromboplastin time) are normal, the specialized tests are required to make an accurate diagnosis. Here, we report a case of acquired FXIII deficiency with severe bleeding symptoms. A 75-year-old man was referred to our hospital because of severe bleeding tendency after a tooth extraction. Laboratory findings showed that routine coagulation studies were normal, but factor XIII (FXIII) activity was low (3%). The presence of FXIII inhibitor was detected with dot blotting studies. Although the bleeding tendency was very severe, it was successfully controlled by infusion of FXIII concentrates combined with immunosuppressive treatment (oral prednisolone). Fibrin cross-linking study showed the significant delay of the γ-chain dimer and α-chain polymer formation. Western blotting revealed the marked decrease in FXIII-A level. The mixing study of FXIII activity measured using amine-incorporation assay showed the incomplete inhibition pattern. There seems to be little agreement as to the treatment strategy of acquired FXIII deficiency. In this patient, the use of FXIII concentrates was very useful in the initial treatment of bleeding symptom. The use of steroids was also effective in increasing FXIII activity without any serious complications.

Selective inducible nitric oxide synthase inhibition attenuates organ dysfunction and elevated endothelin levels in LPS-induced DIC model rats.

We examined the role of nitric oxide (NO) produced by an inducible isoform of NO synthase (iNOS) using N[6]-(iminoethyl)-lysine (L-NIL), a selective iNOS inhibitor, in the rat model of lipopolysaccharide (LPS)-induced disseminated intravascular coagulation (DIC) and investigated changes in organ function, plasma levels of NOX (metabolites of NO) and endothelin. We induced experimental DIC by the sustained infusion of 30 mg kg(-1) LPS for 4 h via the tail vein. We then investigated the effect of L-NIL (6 mg kg(-1), from - 0.5 to 4 h) on LPS-induced DIC. Blood was withdrawn at 4 and 8 h, and all four groups (LPS with or without L-NIL at 4 and 8 h) consisted of eight rats. Three of the animals in the 8-h LPS group died, and we examined blood samples from five rats in this group. None of the other rats died. The LPS-induced elevation of creatinine, alanine aminotransferase, glomerular fibrin deposition and plasminogen activator inhibitor was significantly suppressed by L-NIL coadministration, although L-NIL did not affect the platelet count, fibrinogen concentration or the level of thrombin-antithrombin complex. Moreover, plasma levels of the D-dimer that reflect the lysis of cross-linked fibrin were significantly increased by L-NIL coadministration in the LPS-induced DIC model. Plasma levels of NOX and endothelin were obviously increased by LPS infusion. However, both levels were significantly suppressed in the LPS + L-NIL group, when compared with the LPS group. Although mean arterial pressure (MAP) was significantly decreased between 2 and 8 h compared with the control in the LPS group, this depression was significantly attenuated in the LPS + L-NIL group. Our results suggest that NO induced by iNOS contributes to hypotension (depressed MAP), the progression of hepatic and renal dysfunction, microthrombus deposition and elevated endothelin levels in the rat model of LPS-induced DIC.

Beneficial effect of JTV-803, a new synthetic inhibitor of activated factor X, against both lipopolysaccharide-induced and tissue factor-induced disseminated intravascular coagulation in rat models.

We examined whether JTV-803, a specific activated factor X inhibitor independent of antithrombin III (ATIII), is effective against disseminated intravascular coagulation (DIC) in rat models induced by tissue factor (TF) or lipopolysaccharides (LPS). In male Wistar rats, DIC was induced by a 4 h infusion of thromboplastin (3.75 U/kg) or LPS (50 mg/kg). The rats were given JTV-803 (0.3 or 3 mg/kg, bolus intravenously) (JTV-803 groups) or low molecular weight heparin (LMWH groups) (200 U/kg, bolus intravenously) prior to an injection of TF or LPS. The results showed that JTV-803 was dose-dependently effective against DIC in both TF-induced and LPS-induced rat models. This anti-DIC effect of JTV-803 at higher doses was almost equivalent to that of LMWH in both types of DIC. Plasma ATIII activity was more prominent in the group treated with JTV-803 than in that treated with LMWH. None of rats died in the TF-induced DIC model with or without drug administration. On the contrary, seven of 22 rats died (mortality rate, 31.8%) in the LPS-induced DIC model without drug administration. Although the mortality rate of rats induced with LPS and treated with LMWH was quite high (6/16, 37.5%), none of the LPS-induced rats treated with JTV-803 died. These findings suggested that JTV-803 can treat both TF-induced and LPS-induced DIC models, and that this drug has greater potential in preserving ATIII and in improving the prognosis of DIC.

Decreased plasma activity of antithrombin or protein C is not due to consumption coagulopathy in septic patients with disseminated intravascular coagulation.

We investigated whether depressed plasma antithrombin and protein C activity, considered as a specific finding of disseminated intravascular coagulation (DIC), is due to consumption coagulopathy in septic patients with DIC. An analysis of hemostatic parameters was performed in 139 septic patients (68 with DIC and 71 without DIC). Plasma activity of antithrombin and protein C tended to be significantly decreased in septic patients with DIC but not in those without DIC (p < 0.001). However, when the septic patients were classified into three groups according to the albumin (or choline esterase) level, no significant differences in antithrombin activity or protein C activity were observed between the patients with and without DIC in any of the subgroups. Notably, neither the plasma activity of antithrombin nor protein C was decreased even in septic patients with DIC who had normal plasma levels of albumin (or choline esterase). No significant correlation was observed between plasma levels of thrombin-antithrombin complex (TAT) and antithrombin activity, or between plasma levels of TAT and protein C activity either in septic patients with DIC or without DIC. It is reasonable to conclude that the markedly reduced plasma activity of antithrombin and protein C is not due to consumption coagulopathy in septic patients with DIC.

Depressed plasma activity of plasminogen or alpha2 plasmin inhibitor is not due to consumption coagulopathy in septic patients with disseminated intravascular coagulation.

We have attempted to determine whether depressed plasma plasminogen and alpha2 plasmin inhibitor (or alpha2 antiplasmin) activity is, as a result of consumption coagulopathy, a specific finding of disseminated intravascular coagulation (DIC) in septic patients. The hemostatic parameters of 139 septic patients (68 with DIC and 71 without DIC) were analyzed. Among the group as a whole, plasma activities of plasminogen and alpha2 plasmin inhibitor were significantly depressed in septic patients with DIC relative to those without DIC (P < 0.01 and P < 0.05, respectively). Notably, a significant correlation was observed between plasma levels of albumin and plasminogen activity, as well as between plasma levels of albumin and alpha2 plasmin inhibitor activity both in septic patients with DIC and those without DIC. However, no significant correlation was observed between plasma levels of plasmin-alpha2 plasmin inhibitor complex (PIC) and plasminogen activity, nor between PIC and alpha2 plasmin inhibitor activity either in septic patients with DIC or those without DIC. We concluded that depressed activity of plasminogen or alpha2 plasmin inhibitor is not as a result of consumption coagulopathy, but rather a result of low synthetic function of the liver in septic patients with DIC.

All-trans retinoic acid is partially effective against lipopolysaccharide-induced but not against tissue-factor-induced disseminated intravascular coagulation in rat models.

All-trans retinoic acid (ATRA) has been introduced to the management of acute promyelocytic leukemia (APL) as a differentiation treatment. This drug not only causes complete remission, but also improves disseminated intravascular coagulation (DIC) without adding anticoagulants in APL. We have attempted to determine whether ATRA is effective against DIC in rat models induced by tissue factor (TF) or lipopolysaccharide (LPS), because the anticoagulant effect of ATRA has been considered to induce thrombomodulin upregulation and TF downregulation on endothelial cells as well as on APL cells. In male Wistar rats, DIC was induced by a 4-h infusion of thromboplastin (3.75 U/kg) or lipopolysaccharide (30 mg/kg). The rats were given ATRA orally each day at a dose of 100 mg/kg per day for 1 week before the injection of TF or LPS in ATRA treatment groups, or given low molecular weight heparin (LMWH) 10 min before the injection of TF or LPS (200 U/kg, bolus intravenously) in LMWH treatment groups. No significant changes in hemostatic parameters or markers of organ dysfunction were caused by the ATRA administration, while DIC was significantly improved by LMWH in the TF-induced model. DIC was significantly improved by both ATRA and LMWH in the LPS-induced model. These findings suggested that ATRA was useful for treating DIC only in the LPS-induced model, and that drug efficacy should be carefully assessed because the agents used to induce DIC considerably influenced the outcome.

An enhanced fibrinolysis prevents the development of multiple organ failure in disseminated intravascular coagulation in spite of much activation of blood coagulation.

OBJECTIVES: To investigate the relationship between fibrinolytic enhancement and the development of multiple organ failure (MOF) in disseminated intravascular coagulation (DIC). To detect the useful prognostic index for outcome in DIC. DESIGN: Case-control study. SETTING: A department of internal medicine in a university hospital, a clinical division for diagnosis and treatment, mainly of respiratory diseases, hematologic diseases, DIC, and other diseases requiring critical care medicine. PATIENTS: A total of 69 DIC patients, 31 with MOF. INTERVENTIONS: None. MEASUREMENT AND MAIN RESULTS: The DIC patients with MOF had more elevated levels of tissue plasminogen activator antigen (t-PA) and plasminogen activator inhibitor antigen (PAI), and more depressed levels of plasmin-alpha2 plasmin inhibitor complex (PIC) and fibrin/fibrinogen degradation products than those without MOF, although no significant difference in thrombin-antithrombin complex (TAT) levels was observed. A fibrinolytic enhancement (shown by PIC) was parallel to an activation of blood coagulation (shown by TAT) in DIC patients without MOF, although no such fibrinolytic enhancement was provoked even by much activation of blood coagulation in DIC patients with MOF. Whereas all the patients without MOF were restored from DIC, 14 of 31 patients with MOF were unable to be restored from DIC and died. A significant increase in plasma levels of t-PA and PAI under the condition of sustained hemostatic activation was observed in the patients who died. CONCLUSION: Enhanced fibrinolysis was considered to be the important defense mechanism in preventing the development of MOF in DIC. The increases in plasma levels of t-PA and PAI were poor prognostic markers in DIC. Further careful study may be useful to clarify whether the fibrinolytic therapy is beneficial in clinical DIC patients with MOF.

One missense mutation in the factor X gene causing factor X deficiency--factor X Kanazawa.

We investigated the molecular basis of factor X deficiency in a Japanese patient whose factor X activity and antigen level were 45% and 50% of normal control values, respectively. All exons and intron/exon junctions of the factor X gene were studied using a strategy combining polymerase chain reaction (PCR) amplification and nonradioactive single-strand conformational polymorphism (SSCP) analysis. Exon 5, containing the DNA fragment of the proband, showed aberrant migration by SSCP analysis. All exon-containing DNA fragments amplified by PCR were sequenced, and it was revealed that the proband was a heterozygote for a G --> A substitution in exon 5 of the factor X gene of the proband. This mutation predicts an amino acid replacement of arginine (Arg) for glycine (Gly) at codon 114 in the second EGF-like domain.

[A point mutation at Arg169 (CGG-->TGG) in hereditary protein C deficiency].

We investigated a 56-year-old Japanese man with protein C deficiency, who was referred to our hospital because of venous sinus thrombosis and pulmonary thromboembolism. Protein C (PC) activity and the corresponding antigen level in plasma were 66% and 106% of the normal values, respectively. Both the activity and antigen levels of protein C were reduced by approximately 50% in plasma from the patient's mother. All exons and their flanking intron regions were amplified by PCR from genomic DNA. Sequencing analysis of the PCR fragments revealed that the patient was heterozygous for a C to T substitution at nucleotide position 6218, resulting in a single amino acid substitution of arginine (CGG) by tryptophan (TGG) at codon 169 of the heavy chain. We analyzed the patient, his mother, and normal controls by a Sac II digestion study of exon 7 and found that the patient and his mother had the same C to T point mutation at base 6218. This mutation could have been responsible for the defective activation of the molecule and the resulting thrombotic disorder. The patient is now being treated with warfarin, and so far no further clinical thrombotic episode has occurred.

Elevated levels of free tissue factor pathway inhibitor antigen in cases of disseminated intravascular coagulation caused by various underlying diseases.

Tissue factor pathway inhibitor (TFPI) is primarily synthesized by vascular endothelial cells and is found in vivo in association with endothelial cells, lipoproteins, or in free form. Free TFPI is the most potent and important type, because it is released from endothelial cells following an injection of heparin, or as a result of pathological stimuli. In order to study the role of TFPI in disease, the concentration of free form TFPI was measured in the plasma of 114 patients suffering from disseminated intravascular coagulation (DIC), as the result of several underlying diseases. Plasma antigen levels of free TFPI were significantly higher even in those patients not exhibiting DIC than in normal healthy subjects. These levels were even higher among patients exhibiting DIC, especially those with acute promyelocytic leukemia or cancer, receiving continuous heparin drip infusions. A significant correlation was observed between the plasma antigen levels of free form TFPI and those of fibrin/fibrinogen degradation products, and free form TFPI and plasmin inhibitor complex (r = 0.428, P < 0.0001 and r = 0.329, P < 0.0001, respectively) among 114 DIC patients. There were no significant differences between the plasma levels of free TFPI in DIC patients with or without multiple organ failure. It has been suggested that the plasma levels of free TFPI are closely related to the levels of fibrinolysis occurring in DIC patients, although further study is required to clarify the degree to which TFPI is expressed by endothelial cells during DIC.

Beneficial effect of the active form of vitamin D3 against LPS-induced DIC but not against tissue-factor-induced DIC in rat models.

1Alpha,25-dihydroxyvitamin D3 (active form of vitamin D3; vitamin D3) has been reported to induce the upregulation of thrombomodulin and downregulation of tissue factor (TF) on monocytes. The possibility exists that vitamin D3 prevents the development of disseminated intravascular coagulation (DIC). In particular, monocyte TF production plays an important role in the pathophysiology of DIC in septic patients. We have attempted to determine whether vitamin D3 is effective against DIC in a rat model induced by lipopolysaccharides (LPS) (30 mg/kg, 4 h) or TF (3.75 U/kg, 4 h) using selective hemostatic parameters, markers of organ dysfunction and pathological findings (assessment of glomelular fibrin deposition). Vitamin D3 was administered orally each day at a dose of 2.0 mg/kg/day for 3 days, or low molecular weight heparin (LMWH 200 u/kg; i.v.) was given 10 min before the injection of TF or LPS in each treatment group. Vitamin D3 was effective against DIC in the rat model induced by LPS only, whereas LMWH was effective against DIC in both rat models induced by either TF or LPS. The anti-DIC effect of vitamin D3 was equal to (or more potent than) that of LMWH. The results suggested that vitamin D3 was useful for the treatment of LPS-induced DIC, and that the assessment of a drug's efficacy should be done carefully given the markedly different results obtained according to the agents used to induce DIC.

Elevated plasma levels of free-form of TFPI antigen in hypercholesterolemic patients.

Several studies have previously reported high levels of total tissue factor pathway inhibitor (TFPI) antigen in patients with hypercholesterolemia. The relationship between serum lipid concentrations and total and free-form TFPI antigen in 32 patients with primary type II hypercholesterolemia and 38 age- and gender-matched normolipemic control subjects was studied (Study Group I). Plasma concentrations of total TFPI (tTFPI) antigen, free-form TFPI (fTFPI) antigen, tissue factor antigen, factor VII activity (FVIIc), and prothrombin fragment 1+2 (F1+2) were measured. The median levels of tTFPI, fTFPI, FVIIc, and F1+2 were higher in hyperlipidemic patients compared with those in healthy subjects. The effect of lowering total cholesterol on hypercoagulability in 25 patients with type II hyperlipoproteinemia (Study Group II) were also studied. The median levels of tTFPI, FVIIc, and F1+2 decreased significantly after 6 months of 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase inhibitor therapy in the hypercholesterolemic patients. On the other hand, fTFPI did not change after therapy. Plasma tTFPI was strongly correlated with total cholesterol and low density lipoprotein (LDL)-cholesterol in hyperlipidemic patients. In contrast to the strong correlation between tTFPI and total cholesterol, the correlation between plasma fTFPI and total cholesterol was relatively poor. These results suggest that the activation of the anticoagulant system as well as the activation of the coagulation system may occur in association with hypercholesterolemia. Furthermore, the results of this study may suggest that lowering of total cholesterol in hyperlipidemic patients reduces the thrombin generation in plasma and that down-regulation of LDL does not affect the anticoagulant potency of TFPI in plasma.

Vitamin K administration to elderly patients with osteoporosis induces no hemostatic activation, even in those with suspected vitamin K deficiency.

The administration of menaquinone-4 (MK-4), one of subclasses of vitamin K2, significantly reduces bone loss in postmenopausal osteoporotic women. However, concerns have been raised about whether vitamin K administration alters the hemostatic balance by inducing a thrombotic tendency. We investigated were whether the administration of vitamin K in the form of MK-4 induced a thrombotic tendency in 29 elderly patients with osteoporosis (5 men, 24 women; age range 78.7+/-5.1 years). Patients were administered 45 mg/day (three times a day, 30 min after each meal) of MK-4 for 12 weeks. Blood samples were obtained from the patients at 0, 4 and 12 weeks after the start of MK-4 administration. A number of hemostatic parameters remained stable under the markedly increased plasma levels of MK-4. However, in patients with suspected vitamin K deficiency, whose plasma levels of vitamin K or factor VII were low, vitamin-K-dependent clotting factors such as factor VII and prothrombin were gradually increased after administration of MK-4. No changes in the sensitive molecular markers such as TAT and F1+2, which reflect the amount of thrombin generated in the blood stream, were observed, even in those patients with suspected vitamin K deficiency. These results indicate that MK-4 can be administered safely, with regard to maintaining the hemostatic balance, to osteoporotic patients receiving no anticoagulant therapy.

[Aseptic meningitis induced by high-dose gamma-globulin in an adult woman with antiphospholipid syndrome and chronic idiopathic thrombocytopenic purpura].

It is well known that the administration of high-dose gamma-globulin concentrate is effective in alleviating thrombocytopenia in patients with idiopathic thrombocytopenic purpura (ITP). However, treatment can sometimes induce aseptic meningitis. A 25-year-old Japanese woman with antiphospholipid syndrome and ITP was conditioned with high-dose gamma-globulin concentrate prior to splenectomy. Three days after the initial gamma-globulin course, she experienced severe headache, vomiting, and high-grade fever. Cerebrospinal fluid examination yielded a diagnosis of aseptic meningitis. The patient's clinical symptoms and abnormal cerebrospinal fluid findings disappeared immediately after the discontinuation of gamma-globulin therapy. Cases of aseptic meningitis induced by high-dose gamma-globulin therapy are uncommon in the literature, and most involve children. Moreover, to our knowledge, only 1 Japanese adult ITP case of gamma-globulin-induced aseptic meningitis has been reported to date. Aseptic meningitis may be one of the important adverse effects of the administration of high-dose gamma-globulin concentrate to pediatric as well as adult ITP patients.

Prednisolone inhibits endotoxin-induced disseminated intravascular coagulation and improves mortality in rats: importance of inflammatory cytokine suppression.

In order to determine whether prednisolone has a protective effect against the development of disseminated intravascular coagulation (DIC), we measured the effect of prednisolone on changes in hemostatic parameters and plasma levels of inflammatory cytokines in endotoxin-treated rats. Decreases in platelet count and fibrinogen levels, prolongation of prothrombin time, and increases in the plasma fibrin degradation products and levels of thrombin-antithrombin III (TAT) complex following the administration of endotoxin, all of which are associated with DIC, were significantly suppressed by the administration of prednisolone. Heparin administration significantly suppressed changes in all these parameters except for the decrease in platelet count. The combination of prednisolone and heparin was more effective than either treatment alone. In order to determine whether these effects of prednisolone are correlated with the suppression of inflammatory cytokine production, we examined the relationship between changes in plasma levels of cytokine, the hemostatic parameters listed above, and mortality using a number of intervention regimens designed to alter events of the experimentally induced DIC. Changes in hemostatic parameters associated with DIC following 30 mg/kg per 4 h of endotoxin infusion were significantly suppressed by treatment with 1 mg/kg prednisolone 30 min before beginning endotoxin infusion, followed by administration of 250 U/kg heparin 2 h after the start of endotoxin infusion (prednisolone-endotoxin-heparin regimen). The heparin and prednisolone were administrated subcutaneously. The administration of prednisolone and heparin in the reverse order (i.e. heparin first and prednisolone second: heparin-endotoxin-prednisolone regimen) also suppressed changes in hemostatic parameters, albeit to a smaller degree. Cytokine production was also significantly suppressed by the first treatment, but was not affected by the regimen in which heparin was administered first. Administration of prednisolone alone or heparin alone 30 min before endotoxin significantly reduced the number of renal glomeruli with fibrin thrombi. Plasma levels of creatinine and alanine transferase were reduced only by prednisolone. Increased plasma levels of interleukin-1beta, tissue necrosis factor-alpha and interleukin-6 were suppressed by prednisolone but not by heparin, and there were significant correlations between plasma levels of TAT and cytokines. Prednisolone was more effective than heparin in reducing mortality at 24 h after 100 mg/kg over 4 h of endotoxin infusion (four of 20 versus 15 of 20 deaths for prednisolone and heparin, respectively). These findings suggest that prednisolone inhibits the development of endotoxin-induced DIC and reduces mortality by a different mechanism than heparin, possibly through suppressing the production of inflammatory cytokines. Prednisolone may be efficacious in preventing DIC and multiple organ dysfunction caused by endotoxin.

Prothrombin fragment 1 + 2 measures treatment effect in patients with antiphospholipid syndrome.

Antiphospholipid syndrome (APS) is characterized by recurrent thrombosis. The anticoagulant management of APS thrombosis remains controversial. Few reports on markers of in vivo activation of coagulation have been reported. To determine whether plasma levels of prothrombin fragment 1 + 2 (F1 +2) correlate with thrombotic risk and treatment effect in patients with APS, plasma F1 + 2 levels were followed in 57 patients with this syndrome for more than 2 years. Clinical findings were also observed in these patients. Plasma levels of F1 + 2 in patients with APS were significantly higher when compared with control subjects (p<.05). These results suggest patients with APS are in a hypercoagulable state. Plasma levels of F1 + 2 significantly decreased following treatment with either aspirin, or aspirin plus warfarin (p<.05 and p<.01, respectively). Recurrent thromboses or spontaneous abortions occurred in all eight patients whose plasma levels of F1 + 2 remained higher than 1 nmol/l after treatment with either aspirin alone or no anticoagulants. These patients were subsequently treated with warfarin as well as aspirin, and plasma levels of F1 + 2 decreased to less than 1 nmol/l, with no additional thrombotic events over the remainder of the 2-year follow-up. No fatal bleeding was observed in treated patients. Our results suggest plasma levels of F1 + 2 are useful indicators of successful treatment. It is also suggested that warfarin plus mini-dose aspirin therapy is effective for patients with APS to protect from recurrent thromboses without harmful side effects. Further, prospective cohort studies are needed to substantiate these associations.

Treatment of disseminated intravascular coagulation (DIC) with all-trans retinoic acid in an endotoxin-induced rat model.

Anticoagulant drugs such as heparin are often administered to patients with disseminated intravascular coagulation (DIC) who are also being treated for their underlying disease. The pathophysiology of DIC is so varied that treatment with medications other than anticoagulants may be useful. All-trans retinoic acid (ATRA), which is used for the treatment of acute promyelocytic leukemia (APL), improves DIC in APL. In vitro studies have reported that ATRA caused downregulation of tissue factor and upregulation of thrombomodulin (TM) on endothelial cells as well as APL cells. We examined the effect of ATRA in an endotoxin-induced rat DIC model. DIC was induced in male Wistar rats with a 4-h sustained infusion of endotoxin at a dose of 30 mg/kg. ATRA (20 mg/day) was given every day for 1 week before the injection of endotoxin. ATRA improved the increase in thrombin-antithrombin III (TAT) complex and D-dimer in this model. Fibrin deposition in renal glomeruli was inhibited by ATRA administration, with an increase in the intensity of immunohistochemical TM staining. These findings suggest that ATRA has beneficial effects in the endotoxin-induced rat DIC model. The mechanism may be an upregulation of TM expression on endothelial cells.

Increased plasma levels of free tissue factor pathway inhibitor in patients with Graves' disease.

Tissue factor pathway inhibitor (TFPI) is present in a free-form and in lipoprotein-associated forms in plasma. In this study, the plasma concentrations of total TFPI (tTFPI) and free-form TFPI (fTFPI) were measured in 25 patients with Graves' disease and 25 age-matched healthy subjects, and the relationship between thyroid state and plasma TFPI was examined. Plasma concentrations (median) of tTFPI and fTFPI in Graves' patients who were hyperthyroid were significantly increased compared with Graves' patients who were euthyroid (152 ng/ml versus 124 ng/ml, p < 0.01 and 41.3 ng/ml versus 20.2 ng/ml, p < 0.0001, respectively), and control subjects (152 ng/ml versus 96 ng/ml, p < 0.0001 and 41.3 ng/ml versus 18.7 ng/ml, p < 0.0001, respectively). There was no significant difference in plasma fTFPI concentrations between the euthyroid group and the control group. Plasma fTFPI concentrations correlated closely with thyroid hormone (T3) levels in the patients (r = 0.559, p < 0.005). Serial measurement of individual patients revealed that plasma concentrations of fTFPI and tTFPI were significantly decreased, reaching normal control values upon attainment of euthyroidism. In conclusion, the close correlation between plasma fTFPI and serum thyroid hormone levels suggests that thyroid hormones might influence the synthesis or metabolism of TFPI on the surface of endothelial cells in patients with Graves' disease. This is the first report concerning high concentrations of plasma tTFPI in patients with hyperthyroidism.