Activation of Tissue Factor–Induced Coagulation and Endothelial Cell Dysfunction in Non–Insulin-Dependent Diabetic Patients With Microalbuminuria
Abstract We studied the relationships between albuminuria, tissue factor–induced coagulation, and endothelial cell dysfunction in 67 patients with non–insulin-dependent diabetes mellitus (NIDDM) who were divided into three groups on the basis of their urinary albumin excretion rate (AER). To assess the early phase of tissue factor–induced coagulation, activated factor VII (FVIIa) levels in plasma were measured by a direct fluorogenic assay. As markers of endothelial cell dysfunction, levels of von Willebrand factor (vWF), tissue-type plasminogen activator–plasminogen activator inhibitor–1 (TPA–PAI-1) complex, PAI-1, and tissue factor pathway inhibitor (TFPI) were measured. FVIIa levels were increased in normoalbuminuric NIDDM patients (AER <15 μg/min) when compared with normal control subjects. This FVIIa increase was accompanied by an increase in thrombin–antithrombin III complex (TAT) levels, indicating increased activation of coagulation even in normoalbuminuric patients. In NIDDM patients with microalbuminuria (AER=15-200 μg/min), the FVIIa level, the FVIIa–FVII antigen (Ag) ratio (an indicator of activation of FVII zymogen to FVIIa), and the TAT level were further increased. This group also had higher levels of endothelial cell–derived factors (vWF, TPA–PAI-1 complex, and PAI-1) than the control group. The levels of endothelial cell–derived factors (including TFPI) were highest in the NIDDM patients with overt albuminuria (AER>200 μg/min). In all 67 diabetic patients, AER showed a strong positive correlation with FVIIa (r=.574, P<.0001) and a weakly but still significant correlation with FVIIa-FVII:Ag (r=.365, P=.01), vWF (r=.315, P<.01), and TAT (r=.323, P<.01). FVIIa showed a weaker correlation with vWF (r=.244, P<.05). FVIIa generation concomitant with an increase in AER is probably due to endothelial cell damage. Increased plasma FVIIa levels would produce hypercoagulability in NIDDM patients with microalbuminuria and thus may be a risk for cardiovascular disease.
- non–insulin-dependent diabetes mellitus
- endothelial cell dysfunction
- tissue factor–induced coagulation
- activated factor VII
- Received February 28, 1995.
- Accepted May 30, 1995.
Microalbuminuria, a subclinical increase in the urinary excretion of albumin, is a powerful prognostic indicator of cardiovascular morbidity and mortality in patients with IDDM and NIDDM1 2 as well as in nondiabetic subjects.3 4 5 Several studies have shown that microalbuminuria is associated with other risk factors for cardiovascular disease, such as hypertension, blood lipid abnormalities, left ventricular hypertrophy, and endothelial cell dysfunction.5 6 7 8 9 Hemostatic abnormalities such as hypercoagulability, hypofibrinolysis, and platelet dysfunction have also been described in diabetes mellitus.10 However, the relationship between microalbuminuria and these hemostatic abnormalities remains uncertain.
FVII plays an important role in the initiation of TF-induced coagulation.11 FVII is the single-chain zymogen form of a serine protease that is converted to the two-chain active form (FVIIa) by various coagulation proteases, including factor Xa, factor IXa, factor XIIa, thrombin, and FVIIa. At sites of vascular injury and on exposure of the subendothelium to circulating blood, an integral membrane protein known as TF comes in contact with circulating FVII to form a bimolecular complex. The formation of this complex is widely believed to be the initial event in the extrinsic blood coagulation pathway. TF-producing cells have also been identified in atherosclerotic plaques, where they may play a significant role in thrombosis-associated plaque rupture.12 Healthy individuals have trace levels of circulating FVIIa (0.5% to 1% of the total FVII:Ag level), which may initially activate FVII complexed with cell-surface TF.13 14 15 Thus, an elevated plasma level of FVIIa may indicate hypercoagulability.
An increase of FVIIc has been proposed as an independent cardiovascular risk factor and has been observed in various atherosclerotic diseases, including diabetes mellitus.16 In addition, we recently reported that patients with cardiovascular disease had an increased FVIIa level and proposed that FVIIa may be an independent risk factor for cardiovascular disease.15 Furthermore, our preliminary study indicated that FVIIa levels were also increased in diabetic patients with microalbuminuria.17
The aim of the present study was to clarify the relationship between microalbuminuria, FVII activation, endothelial cell dysfunction, and activation of the coagulation pathway in NIDDM. We found that the generation of FVIIa in NIDDM patients was associated with increases in the urinary AER and endothelial cell damage, resulting in the activation of coagulation. This increased FVIIa level may partly account for the hypercoagulable state in NIDDM patients.
We studied 67 outpatients with NIDDM aged 35 to 79 years (mean age, 67 years) who underwent blood tests more than three times and showed stable glycemic control. All of the patients were ambulatory and had a normal appetite. They and their immediate family members did not have any history of venous thrombosis or pulmonary embolism. Patients with overt renal failure (serum creatinine >1.5 mg/dL and/or BUN >30 mg/dL); liver disease (aspartate aminotransferase or alanine aminotransferase >40 IU/L); intercurrent illnesses except hypertension; and/or a history of coronary artery disease, stroke, congestive heart failure, and malignancy were excluded from this study. ECG was performed on all patients. Asymptomatic patients with an abnormal Q wave on the ECG (Q wave with a duration of 0.04 second or longer) were not excluded from this study. The duration of diabetes was defined as the period from the time that glycosuria was first detected until the time of examination for this study. Smokers were defined as those patients who were currently smoking. Patients with a systolic blood pressure >160 mm Hg and/or diastolic blood pressure >95 mm Hg on three occasions or those receiving antihypertensive therapy were diagnosed as having hypertension. Forty-two normotensive control subjects, matched for age, sex, and body mass index, were also studied. The body mass index was calculated as weight (in kilograms) divided by the square of height (in meters squared).
To avoid the confounding influence of daily physical activity and to facilitate precise specimen collection, we asked the patients to collect urine on two consecutive days between 7 pm and 7 am, and then we determined the mean urinary albumin level. The 67 diabetic subjects were divided into the following three groups on the basis of their AER: a normoalbuminuria group (AER <15 μg/min, n=31), a microalbuminuria group (AER =15 to 200 μg/min, n=25), and an overt albuminuria group (AER >200 μg/min, n=11).
After a minimum 12-hour fast, blood samples for hemostatic determinations were collected into disposable, siliconized, evacuated glass tubes containing 0.1 vol 3.8% trisodium citrate, and blood from the second tube was used for the coagulation assay. The samples were centrifuged at 3000g for 15 minutes at room temperature within 1 hour of collection. Plasma was subsequently separated and stored in several plastic tubes at −80°C until laboratory determinations were performed. The thawed samples were used to determine the levels of TPA–PAI-1 complex, PAI-1, FVII, TAT, and d-dimer.
FVIIa levels were measured by our previously described fluorogenic assay15 using a fluorogenic peptide substrate for thrombin (N-tert-butoxycarbonyl-Val-Pro-Arg-7-amido-4-methylcoumarin, Peptide Institute Inc), congenital human FVII-deficient plasma (George King Bio-Medical), and recombinant, soluble, human TF expressed in yeast and purified.18 Human plasma FVIIa for use as a standard was kindly provided by Dr Tomohiro Nakagaki of the Chemo-Sero-Therapeutic Research Institute (Kumamoto, Japan). The FVIIc level was measured with a chromogenic assay autoanalyzer (Behring Chromotimer, Behringwerke) using a human placental calcified thromboplastin reagent (Chromoquick, Behringwerke) and immunoadsorbed FVII-deficient plasma (Behringwerke AG) as described previously.19 The FVII:Ag level was determined with an ELISA kit (Diagnostica Stago).
Plasma levels of PAI-1 and TPA–PAI-1 complex were determined by using ELISA kits (TDC-88, Teijin Co Ltd)20 and were expressed in nanograms per milliliter. In brief, for measurement of TPA–PAI-1 complex, 100 μL plasma was diluted, mixed, and incubated with a monoclonal anti–PAI-1 antibody–coated polystyrene ball and peroxidase-conjugated monoclonal anti-TPA antibody. After being washed, the ball was immersed in a substrate solution containing H2O2. Then the reaction was stopped by the addition of oxalic acid, and the absorbance of the solution was measured spectrophotometrically. Plasma PAI-1 antigen levels were determined by ascertaining the capacity of plasma to form additional TPA–PAI-1 complexes. Fifty microliters of plasma was preincubated for 30 minutes at 37°C with 50 μL exogenous TPA (500 ng/mL), and the concentration of TPA–PAI-1 complexes was determined. Addition of this amount of TPA resulted in complete saturation of PAI-1.
vWF and d-dimer levels were determined by using ELISA kits (Diagnostica Stago). TAT and TFPI antigen levels were determined by using ELISA kits obtained from Behringwerke and American Diagnostica, respectively. For FVIIc, FVII:Ag, and vWF assays, commercially available pooled plasma (CTS Standard Plasma, Behringwerke AG) was taken as 100%. The FVIIa to FVII:Ag ratio was calculated as an indicator of FVII zymogen activation to FVIIa by taking the mean plasma FVIIa level in young Japanese control subjects (2.1 ng/mL) as 100%.15
Serum total cholesterol and triglyceride levels were determined by using commercial enzyme assay kits (Wako), whereas serum HDL cholesterol was determined by an enzymatic procedure after precipitation with phosphotungstic acid (Wako). Serum glucose was determined by the glucose oxidase method using a commercial enzymatic assay kit (Kanto Chemicals). HbA1c was determined by high-performance liquid chromatography. BUN, creatinine, and uric acid levels were also measured with routine enzyme assay kits. Urinary creatinine was measured by a method based on the Jaffe reaction. The AER was determined by a nephelometric method and expressed as micrograms per minute.
In our laboratory, the interassay coefficient of variation was 4.2% for vWF, 2.3% for TPA–PAI-1 complex, 5.4% for PAI-1, 4.2% for TFPI, 4.2% for FVIIa, 2.8% for FVIIc, 4.4% for FVII:Ag, 3.3% for TAT, 3.0% for d-dimer, and 5.2% for AER.
Data are shown as the mean (with the 95% confidence interval). The distribution of vWF, TPA–PAI-1 complex, PAI-1, TFPI, FVIIa, FVIIc, FVII:Ag, FVIIa-FVII:Ag ratio, TAT, d-dimer, and triglyceride levels was examined, and a logarithmic transformation (to base 10) was performed to reduce the skewness and kurtosis of the distribution curve prior to statistical analysis. The geometric mean of each parameter was determined. One-way ANOVA and unpaired t test were used for comparison of mean values between any two groups. In addition, Pearson’s correlation coefficients were calculated for the different variables. Differences with a value of P<.05 were considered significant.
Table 1⇓ shows the clinical data for the 42 control subjects and the 67 NIDDM patients divided into three groups according to the progression of diabetic nephropathy as indicated by the AER. There were no differences in age and body mass index among the three patient groups, but the duration of diabetes was longer, the prevalence of hypertension higher, and abnormal Q waves on the ECG more common in the albuminuria group. However, there were no significant differences in fasting levels of glucose, HbA1c, cholesterol, triglycerides, HDL cholesterol, uric acid, BUN, and creatinine among the three patient groups.
Table 2⇓ shows plasma levels of FVII-related activity/antigen, endothelial cell–derived molecular markers (vWF, TPA–PAI-1 complex, PAI-1, and TFPI), and markers of coagulation activation (TAT and d-dimer) in the three NIDDM groups and control subjects. The normoalbuminuria group had significantly higher FVIIa, PAI-1, and TAT levels than the control subjects, whereas FVIIc and FVII:Ag were not significantly different. The microalbuminuria group showed a further increase in the FVIIa level accompanied by a significant increase in the FVIIa-FVII:Ag ratio (indicating activation of FVII zymogen to FVIIa). Both increases were significant compared with not only the control subjects but also the normoalbuminuric group. In addition, FVIIc and FVII:Ag levels were significantly increased in the microalbuminuria group compared with the control subjects. Furthermore, the microalbuminuria group had higher levels of vWF, TPA–PAI-1 complex, and PAI-1 but did not show a significant increase in TFPI compared with the control group. The albuminuria group had the highest levels of FVIIa, FVIIc, FVII:Ag, FVIIa-FVII:Ag ratio, TAT, and endothelial cell–derived markers, including TFPI. d-Dimer levels also tended to be increased in the albuminuria group, but the change was not statistically significant.
Fig 1⇓ shows the relation of AER with vWF, FVIIa, FVIIa-FVII:Ag ratio, and TAT in all 67 NIDDM patients. AER showed a strong positive correlation with FVIIa (Fig 1B⇓, r=.574, P<.0001). AER also showed a significant correlation with the FVIIa-FVII:Ag ratio (Fig 1C⇓, r=.365, P<.01) and weaker but still significant correlations with FVIIc (r=.261, P<.05), FVII:Ag (r=.301, P<.05), vWF (Fig 1A⇓, r=.319, P<.01), and TAT (Fig 1D⇓, r=.323, P<.01). In addition, vWF and FVIIa levels showed a slight but significant correlation (Fig 2⇓, r=.244, P<.05).
HbA1c had a strong, positive correlation with FVII:Ag (r=.455, P<.0001), FVIIa (r=.373, P<.01), FVIIc (r=.306, P<.05), and TFPI (r=.380, P<.01) but showed no correlation with the FVIIa-FVII:Ag ratio (not shown). In contrast, fasting glucose levels did not show any correlations with these factors. Total cholesterol showed positive correlations with TFPI (r=.418, P<.001) and FVII:Ag (r=.305, P<.05), whereas triglycerides were correlated with FVII:Ag (r=.253, P<.05) (not shown).
Because 14 of 20 (70%) hypertensive diabetic patients were receiving antihypertensive therapy, the office blood pressure was not adequate to assess the correlation between blood pressure and other parameters. Accordingly, we divided the 67 NIDDM subjects into normotensive and hypertensive groups. In the hypertensive group, AER was higher than in the normotensive group (69 [28 to 150] versus 18 [12 to 29] μg/min, P<.01). The levels of FVIIa, FVIIc, vWF, TPA–PAI-1 complex, and PAI-1 also tended to be increased in the hypertensive group, but the difference was not significant (P<.2).
Table 3⇓ shows various factors in the NIDDM patients with or without abnormal Q waves on the ECG. Those with intercurrent illness and/or a history of coronary artery disease were excluded. The FVIIa, FVIIa-FVII:Ag ratio, and vWF levels were all significantly higher in patients with abnormal Q waves than in those without this finding, but FVIIc and FVII:Ag levels did not differ between these two groups.
The present study has shown that an elevated AER is associated with increased plasma FVIIa levels and endothelial cell damage in patients with NIDDM. Because these pathophysiological changes could result in activation of coagulation systems, the increased FVIIa level may partly account for the hypercoagulability that occurs in NIDDM.
As most previous studies on hypercoagulability did not classify diabetic patients on the basis of AER, it has remained unclear whether or not normoalbuminuric patients have endothelial damage and/or hypercoagulability. Several recent articles have indicated that the levels of vWF,9 thrombomodulin (a marker of endothelial damage),21 and factor XIa–α1-antitrypsin complex (a marker of activation of the contact phase of coagulation)22 are not increased in normoalbuminuric NIDDM patients, suggesting that systemic endothelial damage and hypercoagulability do not pose a problem for these patients. In contrast, our data indicated that FVIIa, PAI-1, and TAT levels were significantly increased in normoalbuminuric patients compared with healthy control subjects. Thus, hemostatic abnormalities were present even in our NIDDM patients with normoalbuminuria. However, the FVIIc and FVII:Ag levels of this group were not significantly different from those of control subjects. Thus, measurement of FVIIc seems to be less sensitive for assessing FVII hyperactivity than does measurement of FVIIa.
FVIIc has previously been shown to be increased in diabetic subjects,23 24 and a positive correlation between FVIIc and glucose levels has been observed in large population studies.25 26 Some authors have reported significantly higher FVIIc levels in microalbuminuric than normoalbuminuric subjects,27 28 but others have not found this difference.29 30 Our study indicated that FVIIc as well as FVII:Ag levels were slightly but significantly different in the microalbuminuria group compared with those from the healthy control subjects. These discrepancies may be partly attributable to the assays used for FVIIc, because the sensitivity of an FVIIc assay for detecting FVIIa is markedly influenced by the thromboplastin and FVII-deficient plasma preparations that are employed.31 32
To assess endothelial function, we measured plasma levels of vWF as well as other endothelial cell–derived factors, including TPA–PAI-1 complex, PAI-1, and TFPI. vWF is a glycoprotein secreted by the vascular endothelium,33 and an increased plasma vWF level is now accepted as an indicator of systemic endothelial dysfunction in NIDDM patients, with the endothelial damage suggested by the increase in this factor possibly explaining the linkage between microalbuminuria and cardiovascular death in these patients.8 9 Other endothelial cell–derived factors are not well characterized in terms of their release from the endothelium as a result of vascular dysfunction. In the present study, we found that only the PAI-1 level was statistically increased in the normoalbuminuria group compared with the healthy control group, whereas the levels of PAI-1, vWF, and TPA–PAI-1 complex were increased in the microalbuminuria group, suggesting a hyperfibrinolytic state and endothelial cell damage as described previously.8 9 10 21 An increase in TFPI was detected in the overt albuminuria group.
The ELISA used in this study enabled us to specifically measure the plasma level of active PAI-1 antigen that could form complexes with TPA.20 The mechanism of PAI-1 release is quite different from that for vWF. vWF can be released from an organelle of endothelial cells, known as the Weibel-Palade body, by various stimuli.34 PAI-1 is an acute-phase reactant that can be detected when lipopolysaccharide or cytokines are added to cultured endothelial cells; it is also found in patients with sepsis, trauma, or a postoperative state.35 An increase of PAI-1 levels has been reported in NIDDM, although the underlying mechanism remains controversial. In vitro studies have suggested that insulin, its precursors, or hyperglycemia itself stimulates PAI-1 synthesis by hepatocytes or endothelial cells.36 37 The different release mechanisms of these substances would explain the different behavior of vWF and PAI-1 in NIDDM patients with normoalbuminuria.
In NIDDM patients with microalbuminuria or overt albuminuria, markedly elevated FVIIa levels were observed, indicating an increase in the early phase of TF-induced coagulation. It is noteworthy that the increase in FVIIa was far higher than that of FVIIc or FVII:Ag, indicating that activation of FVII was markedly enhanced. Factor Xa, α-thrombin, and possibly FVIIa are thought to catalyze the initial activation of FVII in complex with a cell-surface procoagulant protein, TF, under various pathological conditions.11 There were significant correlations between AER and FVIIa level (Fig 1B⇑, r=.574), vWF and AER (Fig 1A⇑, r=.319), or vWF and FVIIa (Fig 2⇑, r=.244). Elevated levels of plasma vWF have been reported to reflect systemic endothelial cell damage and to be related to AER in diabetes.8 9 At the present time, it remains uncertain whether FVIIa generation in diabetes is due to systemic endothelial cell damage or localized renal endothelial damage in a TF-dependent manner. To address this issue, TF expression in the kidneys of diabetic patients must be studied by an in situ hybridization method or immunohistochemistry technique.
The albuminuria group had the highest levels of endothelial cell–derived factors, FVII, and TAT among the three patient groups. In our previous study of patients with arterial cardiovascular disease, a mean FVIIa level of 3.3 ng/mL was obtained.15 Thus, the mean FVIIa levels of 3.9 ng/mL in the microalbuminuria group and of 4.7 ng/mL in the overt albuminuria group were both much higher than those that are typically found in healthy young subjects (2.1 ng/mL) or in patients with cardiovascular disease. Such findings could explain the worse clinical outcome in these groups. In addition, FVIIc and FVII:Ag levels were also significantly increased along with AER, suggesting that proteinuria stimulates the hepatic synthesis of FVII. Concerning coagulation inhibitors, the levels of antithrombin III, protein C, and protein S have been reported to be decreased in albuminuric patients, probably due to urinary loss.38 On the contrary, TFPI levels were increased in the overt albuminuric group. Although the precise mechanism that underlies increases in TFPI levels in the overt albuminuria group remains uncertain, more extensive vascular damage might release TFPI from endothelial cell glycosaminoglycans.
In the present study, we excluded those patients with a history of coronary artery disease, but we still found nine asymptomatic patients with abnormal Q waves on the ECG. In these patients with myocardial ischemia, vWF, FVIIa, and the FVIIa-FVII:Ag ratio were higher, but FVII:Ag levels did not differ when compared with those with no abnormal ECG findings (Table 3⇑). Thus, endothelial cell damage and FVII activation were present in these patients.
In conclusion, generation of FVIIa increases along with AER in NIDDM patients. This FVIIa generation is probably due to endothelial cell damage, with subsequent activation of coagulation, and might explain the higher risk of cardiovascular disease in NIDDM patients with microalbuminuria. To confirm the predictive value of increased FVIIa levels in these patients for the hypercoagulable state and subsequent cardiovascular episodes, further studies are needed in a prospective cohort setting.
Selected Abbreviations and Acronyms
|AER||=||albumin excretion rate|
|BUN||=||blood urea nitrogen|
|ELISA||=||enzyme-linked immunosorbent assay|
|FVIIa||=||activated factor VII|
|FVII:Ag||=||factor VII antigen|
|FVIIc||=||factor VII coagulant activity|
|(N)IDDM||=||(non–)insulin-dependent diabetes mellitus|
|PAI-1||=||plasminogen activator inhibitor–1|
|TAT||=||thrombin–antithrombin III complex|
|TFPI||=||tissue factor pathway inhibitor|
|TPA||=||tissue-type plasminogen activator|
|vWF||=||von Willebrand factor|
This study was supported in part by grants-in-aid from the Foundation for the Development of the Community, Tochigi; from the Ministry of Education, Science, and Culture of Japan; from Osaka Gas Group Welfare Foundation, Osaka; and from Special Coordination Funds for Promoting Science and Technology (Encouragement System of COE), the Science and Technology Agency of Japan. We appreciate Dr Hisao Kato at National Cardiovascular Center Research Institute for his encouragement and helpful discussion. We thank Dr Tomohiro Nakagaki at the Chemo-Sero-Therapeutic Research Institute, Kumamoto, Japan, for a gift of purified human plasma FVIIa.
Damsgaard EM, Frøland A, Jørgensen OD, Mogensen CE. Microalbuminuria as predictor of increased mortality in elderly people. Br Med J. 1990;300:297-300.
Haffner SM, Stern MP, Gruber MKK, Hazuda HP, Mitchell BD, Patterson JK. Microalbuminuria: potential marker for increased cardiovascular risk factors in nondiabetic subjects? Arteriosclerosis. 1990;10:727-731.
Pedrinelli R, Di Bello V, Catapano G, Talarico L, Materazzi F, Santoro G, Giusti C, Mosca F, Melillo E, Ferrari M. Microalbuminuria is a marker of left ventricular hypertrophy but not hyperinsulinemia in nondiabetic atherosclerotic patients. Arterioscler Thromb. 1993;13:900-906.
Jensen T. Increased plasma concentration of von Willebrand factor in insulin dependent diabetics with incipient nephropathy. Br Med J. 1989;298:27-28.
Sandeman DD, Tooke JE. Diabetic vascular disease. In: Bloom AL, Forbes CD, Thomas DP, Tuddenham EGD, eds. Haemostasis and Thrombosis. London, England: Churchill Livingstone Inc; 1994:1291-1299.
Wilcox JN, Smith KM, Schwartz SM, Gordon D. Localization of tissue factor in the normal vessel wall and in the atherosclerotic plaque. Proc Natl Acad Sci U S A. 1989;86:2839-2843.
Morrissey JH, Macik BG, Neuenschwander PF, Comp PC. Quantitation of activated factor VII levels in plasma using a tissue factor mutant selectively deficient in promoting factor VII activation. Blood. 1993;81:734-744.
Wildgoose P, Nemerson Y, Hansen LL, Nielsen FE, Glazer S, Hedner U. Measurement of basal levels of factor VIIa in hemophilia A and B patients. Blood. 1992;80:25-28.
Kario K, Miyata T, Sakata T, Matsuo T, Kato H. Fluorogenic assay of activated factor VII: plasma factor VIIa levels in relation to arterial cardiovascular diseases in Japanese. Arterioscler Thromb. 1994;14:265-274.
Meade TW. The epidemiology of atheroma, thrombosis and ischaemic heart disease. In: Bloom AL, Forbes CD, Thomas DP, Tuddenham EGD, eds. Haemostasis and Thrombosis. London, England: Churchill Livingstone Inc; 1994:1199-1227.
Shigematsu Y, Miyata T, Higashi S, Miki T, Sadler JE, Iwanaga S. Expression of human soluble tissue factor in yeast and enzymatic properties of its complex with factor VIIa. J Biol Chem. 1992;267:21329-21337.
Sakata Y, Murakami T, Noro K, Mori K, Matsuda T. The specific activity of plasminogen activator inhibitor-1 in disseminated intravascular coagulation with acute promyelocytic leukemia. Blood. 1991;77:1949-1957.
Iwashima Y, Sato T, Watanabe K, Ooshima E, Hiraishi S, Ishii H, Kazama M, Makino I. Elevation of plasma thrombomodulin level in diabetic patients with early diabetic nephropathy. Diabetes. 1990;39:983-988.
Murakami T, Komiyama Y, Egawa H, Murata K. Elevation of factor XIa-α1-antitrypsin complex levels in NIDDM patients with diabetic nephropathy. Diabetes. 1993;42:233-238.
Fuller JH, Keen H, Jarret RJ, Omer T, Meade TW, Chakrabarti R, North WRS, Stirling Y. Haemostatic variables associated with diabetes mellitus and its complications. Br Med J. 1979;2:964-966.
Balleisen L, Assmann G, Bailey J, Epping P-H, Schulte H, van de Loo J. Epidemiological study on factor VII, factor VIII and fibrinogen in an industrial population, II: baseline data on the relation of blood pressure, blood glucose, uric acid, and lipid fractions. Thromb Haemost. 1985;54:721-723.
Gruden G, Cavallo-Perin P, Bazzan M, Stella S, Vuolo A, Pagano G. PAI-1 and factor VII activity are higher in IDDM patients with microalbuminuria. Diabetes. 1994;43:426-429.
Jones SL, Close CF, Mattock MB, Jarrett RJ, Keen H, Viberti GC. Plasma lipid and coagulation factor concentrations in insulin dependent diabetics with microalbuminuria. Br Med J. 1989;298:487-490.
Jaffe EA, Hoyer LW, Nachman RL. Synthesis of von Willebrand factor by cultured human endothelial cells. Proc Natl Acad Sci U S A. 1974;71:1906-1909.
Sadler EJ. von Willebrand factor. J Biol Chem. 1991;266:22777-22780.
Nordt TK, Schneider DJ, Sobel BE. Augmentation of the synthesis of plasminogen activator type-1 by precursors of insulin: a potential risk factor for vascular disease. Circulation. 1994;89:321-330.
Nordt TK, Klassen KJ, Schneider DJ, Sobel BE. Augmentation of synthesis of plasminogen activator inhibitor type-1 in arterial endothelial cells by glucose and its implications for local fibrinolysis. Arterioscler Thromb. 1993;13:1822-1828.