Effects of an Oral Antidiabetic Drug on the Fibrinolytic Insulin-Treated Diabetic Patients Jdrgen Gram, Jbrgen Jespersen,
System of Blood in
and Aage Kold
Selected variables of the fibrinolytic system were assessed in 23 men with insulin-treated diabetes with no measurable pancreatic &cell function. Gliclazide, a second-generation sulphonylurea drug, was administered to the patients over a period of 6 months in daily doses of 160 mg or 240 mg, and blood samples were obtained before, during, and after treatment. Determined by global assays, the drug did not significantly change plasminogen activator activities in euglobulins. Measurements of specific components of the system of fibrinolysis showed a marginal increase during administration of gliclaside of tissue-type plasminogen-activator antigen and prekallikrein activity in plasma, whereas the activities in euglobulins of the intrinsic plasminogen proactivators remained nearly the same during the study. Levels in plasma and euglobulin of Cl-inactivator antigen and in plasma of factor XII antigen and t-PA inhibition capacity remained constant throughout the study. There were no changes of the increase in concentration of t-PA activity and t-PA antigen following venous occlusion. The metabolic state remained the same during the whole study. It is concluded that gliclazide non-insulin-dependent extrametabolic effects on the extrinsic (t-PA) and intrinsic induces small, but significant, (prekallikrein) system of fibrinolysis. Whether these changes are of physiological importance remains to be demonstrated. m 1988 by Grune & Stratton, inc.
ESPITE numerous clinical and laboratory studies, the etiology and pathogenesis of the vascular manifestations of late diabetes remain enigmatic. Several authors have suggested that deposition of fibrin in the walls of the vessels of the microcirculation might constitute a factor of pathogenetic significance. I-3 This has caused many to look for defects in the hemostatic system of diabetic patients with particular emphasis on the possible existence of an impaired fibrinolytic system. Implicit in such a concept is the possibility that an improvement of a depressed fibrinolytic activity in blood might help to repair a damaged microcirculation and delay or prevent the vascular complications of late diabetes. Gliclazide, a second-generation sulphonylurea drug, has been reported to increase blood fibrinolytic activity in animals and in patients with type II diabetes.4-7 However, since the clinical studies dealt with non-insulin-dependent diabetics, the reported effect might have resulted, in part, from the insulin-mediated metabolic effect of the drug, and might not have been caused by a direct effect on the fibrinolytic system. To resolve this uncertainty, it is necessary to determine the effects of the drug on diabetic patients already receiving treatment with insulin. As an introduction to a study of the effects of gliclazide on type II diabetics, we have therefore applied the drug to a well-controlled group of patients with type 1 diabetes (no &cell function) receiving treatment with insulin.
D
weight or had a systolic BP above 150 mm Hg. Patients receiving drugs with known antithrombotic properties (vitamin K-antagonists, acetylsalicylic acid, etc) were not included in the study. The total study period was 8 months, during which the patients were randomly divided into two groups to receive either one daily oral dose of 160 mg (n = 11) or of 240 mg (n = 12) gliclazide (Diamicron, Les Laboratoires Servier, Nevilly-sur-Seine, France), each for a period of 6 months. At seven points in the study, samples of blood (4.5 mL in 0.5 mL 0.13 mol/L trisodium citrate; 3.0 mL in 0.06 mL 0.17 mol/L tri-potassium EDTA) were collected in the morning from fasting patients: (1) 1 month before treatment, (2) immediately prior to administration of gliclazide, (3) after 1 month of treatment, (4) after 2 months, (5) after 4 months, (6) after 6 months, and (7) 1 month after cessation of treatment. Immediately prior to administration of gliclazide, and again after 6 months of treatment, blood samples also were obtained after venous occlusion (10 mm Hg above the diastolic BP for ten minutes). Blood sampling and the preparation and handling of plasma samples followed standardized procedures.8 Assays
Patients
Metabolic control. The following variables were determined: serum glucose by a dehydrogenase method on a Technicon RA- 1000 analyzer (Technicon Instruments Corporation, New York), hemoglobin A,, (HbA,,) by isoelectric focusing: apolipoprotein (APO) A and APO B in serum on M-Partigen plates (Behringwerke, Marburg, FRG), and triglyceride (TGLY) and total cholesterol (CHOL) in serum by enzymatic methods using available kits (Boehringer, Mannheim, FRG). Acute phase reactants. Selected for determination were three complementary reactants! fibrinogen (FIBR) by a modified thrombin time assay,” plasminogen (PLG) in plasma by an automated
Volunteering in the study were 24 informed, male patients with type I diabetes (median age 47 years, range 23 to 58; median duration 15 years, range 6 to 27) with signs of microvascular damage in the form of background retinopathy, but without proliferative characteristics. One patient was withdrawn from the study. Women were excluded in order to avoid fluctuations in the determined variables caused by the menstrual cycle. A defective pancreatic @cell function was verified by a failing C-peptide release after the intravenous injection of 1 mg glucagon (Novo Industri, Copenhagen). All patients were receiving well-regulated treatment with insulin twice daily (human type, Novo Industri, Copenhagen). None of the patients exceeded by more than 10% the normal average
From the Section of Coagulation and Fibrinolysis. Department of Clinical Chemistry, and the Department of Internal Medicine, Ribe County Hospital in Esbjerg; Section of Thrombosis Research. South Jutland University Centre, Esbjerg, Denmark. Supported by Les Laboratoires Servier. Armedic, Denmark. Address reprint requests to J$rgen Gram, MD, Section of Coagulation and Fibrinolysis, Department of Clinical Chemistry, Ribe County Hospital in Esbjerg, DK-6700 Esbjerg, Denmark. o 1988 by Grune & Stratton, Inc. 0026-0495/88/3710-0007$03.00/O
MATERIALS
AND METHODS
Metabolism, Vol37, No 10 (October), 1988: pp 937-943
937
938
GRAM, JESPERSEN, AND KOLD
method using streptokinase and modified to correct for the overestimation in pathologic plasma samples,” and C-reactive protein (CRP) in serum by a turbidimetric assay (Orion Diagnostica, Helsinki, Finland). Fibrinolysis assays. Plasma samples for fibrinolysis assays were kept at -80°C, and the euglobulin specimens were prepared immediately prior to assay. Each assay was run during one session on all samples in order to eliminate interassay variations. Assayed were the main components of the fibrinolytic system depicted in Fig 1. Euglobulin activity (globof assay). Normal euglobulin fractions (NEF) were prepared from plasma in 1:lO dilution (1 vol plasma and 9 vol ice-cold distilled water) at pH 5.9.” Activity was determined on standardized, plasminogen-rich fibrin plates as the mean area of the lysed zones (diameter products, mm’, triplicates) produced by 30 pL of the active solution after 17 hours at 37OCin a moist incubator. Extrinsic (cellular derived) system ofjibrinolysis. The concentration of extrinsic (tissue-type, t-PA) PLG activator was measured in the NEF by a spectrophotometric assay.14 The t-PA antigen in plasma was determined by an enzyme immunoassay (Biopool St-PA, Biopool, Umea, Sweden). Plasminogen activator inhibition (PA-I) in plasma was determined with purified two-chain human melanoma t-PA by a spectrophotometric assay” modified from the titration assay of Verheijen et al.16 Intrinsic (humorul derived) system offibrinolysis. The intrinsic PLG activator system (cf Fig 1) comprises a factor XII-dependent and a urokinase-related, factor XII-independent pathway.” The normal euglobulin precipitate contains only a fraction of the potentially available intrinsic PLG activators.‘3,‘8,‘9 In the presence of high-molecular-weight anionic compounds, in particular highmolecular-weight dextran sulphate (M, 500,000). additional amounts of PLG activator are recovered in the euglobulin precipitates.‘9,20Therefore, for determination of intrinsic PLG activators on fibrin plates, a dextran sulphate euglobulin fraction (DEF) was prepared by dilution of 1 vol plasma with 8 vol ice-cold distilled water followed by addition of 1 vol of a solution of dextran sulphate 500,000 (100 mg/L, Pharmacia, Uppsala, Sweden) and adjustment of pH to 5.9.” To eliminate the inhibitory effect of the Clinactivator co-precipitated with the euglobulin fraction, a solution of sodium flufenamate (flufenamic acid, Aldrich Europe, Beerse, Belgium) was added to the euglobulin solution to a final concentration
yy;T_x;; PREKALLIKREIN
of 2 mmol/L.” To quench the urokinase-related intrinsic activity, goat antiserum against high-molecular-weight urokinase (IgG) was added to the DEF solution (containing flufenamate) before application on fibrin plates.” The titer of the antiserum solution was adjusted to quench the total urokinase-related activity in the plasma. The t-PA present in the plasma is also precipitated in the DEF. The level of t-PA in morning samples from resting subjects is normally low and contributes insignificantly to the activity of the dextran sulphate euglobulin fraction. In our laboratory, the median t-PA activity in a group of healthy volunteers (n = 43) was only 26 mm’. Other hemostatic assays. Rocket immunoelectrophoresis was used to determine the level of factor XII in plasma (antiserum from Nordic, Tilburg, The Netherlands) and of the Cl-inactivator in plasma and in NEF (antiserum from Dakopatts, Glostrup, Denmark). Prekallikrein in plasma was assayed spectrophotometrically on a Cobas Bio centrifugal analyzer using a commercially available prekallikrein activator (KabiVitrum, Stockholm). Statistical analysis. Fluctuations during the seven periods of measurement were evaluated by nonparametric ANOVA using the Friedman Xz test corrected for ties. Differences between two measurements (venous occlusion tests) were estimated by the Wilcoxon signed rank sum test. RESULTS
The daily insulin requirements of the two groups did not change significantly during the study. Patients receiving 160 mg gliclazide were administered a daily dose (median values and 95% interpercentile ranges of medians) of 54 (32 to 68) IU insulin prior to the study and 54 (30 to 66) IU insulin by the end of the study. The corresponding values for patients receiving 240 mg were 39 (28 to 60) IU insulin before the study and 44 (24 to 60) IU insulin by the end. Metabolic
Status
The levels of fasting glucose showed variations in both groups during the period of treatment and study, although not reaching statistical significance (Fig 2). The long-term metabolic state of the patients (as reflected in the recorded levels of HbA,,, lipoproteins, CHOL, and TGLY) was not changed by the treatment in either of the two groups (160 mg and 240 mg gliclazide, respectively; Fig 2).
Iggy -
KALLIKREIN’
\
Acute Phase Reactants The levels of PLG and FIBR fluctuated insignificantly during the study (Fig 3). All but one patient had levels of CRP 5 0.10 g/L, ie, normal values (not shown). The sole deviating patient had a marked temporary elevation that suggested an acute disease state, wherefore he was withdrawn from the study, thereby leaving a group of 11 patients consuming one daily dose of 160 mg gliclazide.
Pt ASMbKXEN
Fibrinolysis FlRRlNOtYSlS
Fig 1. The activation of blood fibrinolysis by (a) an extrinsic t-PA pathway, (b) a factor XII-dependent plasminogen proactivator pathway, and (c) a urokinese-related plasminogen proactivator pathway. Abbreviations: Cl-INA, Cl-esterase inactivator; PROUK, prourokinase; UK, urokinase: PAI, plasminogen activator inhibitor.
Variables
The fibrinolytic activities of the regular plasma euglobulins (NEF, global assay) did not change significantly in the patients receiving 160 mg gliclazide, whereas there was a small, but statistically significant, fall of activity in the group receiving 240 mg (P -C .05, X2 = 20.5, df = 6) (Fig 4A). Since the regular euglobulin in addition to its t-PA content contains a part of the potentially available activators of the
FIBRINOLYSISAND ORAL ANTIDIABETIC DRUGS
939
HB A,C
0.0
4.p”
1
-
APO A
PLG
mmoltl
TGLY
0.0
1
I t
2
I 3
4
5
6
7
Fig 3. Median levels in plasma of FIBR and PLO in the two groups of diabetics. Levels of plasminogen expressed as percentages of the concentration in pooled normal plasma.
Fig 2. Median levels in serum of selected metabolic variables in two groups of type I diabetics given one daily dose of gliclazide 160 mg (A: n = 11) or 240 mg (0: n = 12). respectively. Glucose (Glu), HLA,,. serum APO A, APO B. CHOL. and TGLY. Abscissa: periods of blood sampling. Treatment period is indicated by open area. Brackets indicate 95% interpercentile intervals of the medians. Periods are slightly displaced as indicated on the abscissa to facilitate identification and improve legibility.
intrinsic system as well as co-precipitated inhibitors*‘,** affecting the measured fibrinolytic activity, it is necessary to apply selective methods of assay. The effect of co-precipitated inhibitors can be eliminated by addition of sodium flufenamate (2 mmol/L) to the euglobulin solution.‘3.2’ With added flufenamate (NEF + F), the median activities at each period of collection show a uniform pattern in both groups with no statistically significant changes (Fig 4B). The median levels of Cl-inactivator in plasma and in NEF remained constant (cf Fig 5), thereby excluding the Clinactivator as a possible contributor to fluctuations in measured euglobulin activities. Analysis of the specific components of the extrinsic system (Tables I and 2) did not show statistically significant changes of active t-PA in euglobulins, or of PA-I measured in plasma. The median concentration in plasma of t-PA antigen increased in both groups of patients after a few months of treatment, reaching statistical significance in patients receiving 160 mg (P < ,005, X2 = 18.9, df = 6) and near significance at the 5% level in patients receiving 240 mg (.05 < P -=z.lO, X2 = 11.l, df = 6). Determination of the increase in concentration of active t-PA after venous occlusion, a measure of the fibrinolytic potential, and calculation of the median individual increases yielded the values at the
two sampling periods shown in Table 2. There were no statistically significant differences between the t-PA increases in the two treated groups or throughout the period of study. The levels of t-PA antigen (Table 2) paralleled the concentration of active t-PA, indicating that the rise in specific activity (IU/mg) of t-PA after venous occlusion did not change. The recorded t-PA activities, the concentrations of t-PA antigen, and the calculated increases were corrected for changes in the plasma volume (hemoconcentration) during the venous occlusion test by determining for each individual the extinction (E,,,) of diluted plasma samples before and after the occlusion. Measurements of the intrinsic fibrinolytic system are shown in Figs 5 and 6. The activity of the dextran sulphate precipitated euglobulins in the presence of flufenamate (DEF) is a measure of the total level of potentially active
mm2
A
800
Fig 4. Median levels of fibrinolytic activity (mm’, logarithmic scale) in (A) solutions of the NEF alone and (B) after addition of flufenamate prepared from samples of plasma obtained in the morning from two groups of diabetics receiving 160 mg (Al or 240 mg (0) gliclazide. respectively.
GRAM, JESPERSEN, AND KOLD
%
F XII
%
krein, a zymogen necessary for optimal activation of the factor XII-dependent pathway of fibrinolysis, increased in concentration, being most pronounced in the patients receiving 160 mg gliclazide (P < .OOl; X2 = 31.3; df = 6) and slightly less pronounced in the group receiving 240 mg (P < .05; X2 = 13.0; df = 6) (Fig 5). In both groups of patients, the plasma concentrations of factor XII remained unchanged throughout the study.
PREK
/
I
DISCUSSION P-ClINA
%
E-ClINA
Attempts to determine and evaluate specific effects of oral antidiabetic agents on the fibrinolytic system in patients with diabetes relate to a rather complex physiological model. Changes in the metabolic state of these non-insulin-dependent patients that are possibly induced by drug-released insulin might influence the fibrinolytic system,24-27thereby compromising a distinction between a specific action and an indirect, metabolic effect of the drug. To eliminate a possible insulin-induced drug effect on the metabolic state, we decided to administer the drug to patients without significant pancreatic P-cell function (type I diabetics) receiving treatment with insulin. This was done in order to be better able to evaluate the effects of the drug on the group of patients for whom it is intended (type II diabetics). The study was designed so that each patient served as his own control, and differences were estimated on the basis of individual deviations. We found such a study design most appropriate because of the large interindividual variance in hemostatic variables in patients with diabetes (cf 28). We found the levels of HbA,, and the average serum glucose levels, as well as additional metabolic variables, to be almost constant during the study (Fig 2). Some variables (FIBR, PLG, CRP) among the so-called acute-phase reactants2’.” were serially monitored in order to be able to exclude any concomitant deviation associated with an inflammatory reaction (Fig 3). One patient showing a temporary, but significant, increase in CRP was withdrawn from the study. The insulin requirements did not change during
Fig 5. Median concentrations of blood coagulation factor XII (F Xii) and prekallikrein (PREK) in plasma, and of Cl-inactivator in plasma (P-Cl-INA) and in normal euglobulins (E-Cl-INA) in the two groups of diabetics. Results expressed as percentages of the concentration in normal pooled plasma.
plasminogen activators in plasma. In both groups of patients, the DEF activities remained constant throughout the whole period of study (Fig 6, upper curves). Control experiments confirmed that the contribution of extrinsic activator (t-PA) to the total activator activity in DEF is negligible (not shown), making it unnecessary to eliminate this insignificant contribution when identifying the intrinsic activators responsible for the DEF activity. ” Measurements of the specific components of the intrinsic system indicate nearly unchanged contributions from the urokinase-related pathway and from the factor XII-dependent pathway of fibrinolysis during the period of study (Fig 6), whereas prekalli-
Table 1. Fibrinolysis Variables. Median Concentrations l of t-PA in Euglobulin, PA-I in Plasma, and t-PA Antigen in Plasma in Type Diabetics Given One Daily Dose of Gliclaxide of Either ? SO mg (n - 11) or 240 mg (n - 12) for Six Months (Periods 2 to 6) ControlPeriod 1
ControlPeriod
Treatment Period 2
3
4
5
I
6
7
t-PA in euglobulin (IU/mL) 160 mg 240 mg
0.02
0.01
0.01
0.04
0.01
0.01
0.0 1
(0.00 to 0.14)
(0.00 to 0.17)
(0.00 to 0.16)
(0.01 to 0.12)
0.04 (0.00 to 0.32)
(0.00 to 0.05) 0.03
(0.00 to 0.05) 0.02
(0.00
(0.00 to 0.09)
(0.00 to 0.14)
0.08
to 0.26)
0.08
0.02
(0.00 to 0.07) 0.05
(0.00 to 0.21)
(0.00 to 0.33)
(0.00 to 0.26)
PA inhibitor in plasma W/mL) 160 mg 240 mg
6.2 (1.9 to 52.9) 4.3 (3.8 to 8.1)
5.8 (2.5 to 44.0) 4.9 (3.6 to 7.8)
6.3 (3.2 to 31.2) 6.1 (3.5 to 7.7)
5.2 (2.4 to 21.0) 6.5
7.4 (2.6 to 27.6) 5.5
(3.6 to 8.4)
(3.6 to 7.7)
6.0 (4.3 to 14.8)
(4.4 to 11.7)
7.2 (3.3 to 64.0) 6.1 (4.3 to 10.0)
5.2 (3.2 to 18.6) 6.2 (4.3 to 8.2)
t-PA antigen in plasma (ng/mL) 160 mg 240 mg
6.3 (3.8 to 10.0) 4.4 (3.1 to 5.9)
6.7 (4.3 to 14.4) 4.6 (2.7 to 8.2)
5.7 (2.8 to 9.1) 3.9 (3.1 to 7.2)
*Ninety-five percent interpercentile interval of the median is given in parentheses.
3.8 (2.7 to 5.5)
7.8 5.1 (3.7 to 8.5)
7.7 (4.2 to 14.1) 4.8 (3.1 to 7.1)
7.7 (3.2 to 11.1) 4.0 (2.8 to 5.7)
941
FIBRINOLYSIS AND ORAL ANTIDIABETIC DRUGS
Table 2. Median Values of Increments* of t-PA Concentration and Antigen Concentration After Venous Occlusion in Two Groups of Type
I Diabetics
BeforeGliclluide 160 mg
240 mg
1.6
2.2
0.6
1.52
(0.0 to 5.6)
(0.0 to 7.8)
(0.0 to 7.9)
1.3
t-PA antigen concentration in plasma (ng/mL)
16Omg
(0.0 to 16.1)
t-PA concentration in euglobulin WmL)
With Gliclazide 240 mg
(O.Oto 16.1)
1.8 (O.Oto6.1)
6.8 (2.3 to 16.6)
2.8 (0.0 to 9.4)
Venous occlusion and collection of blood samples was immediately prior to administration of gliclazide and following 6 months of treatment. *Values were corrected for hemoconcentration.
the study, and none of the tested nonspecific metabolic variables showed in any of the groups’ fluctuations, which could be related to the administration of gliclazide. In the study of plasma fibrinolytic activity, a euglobulin solution is usually prepared followed by assessment of its activity, eg, by the fibrin plate method.13 Three distinct types of PLG activator in plasma can be identified: an extrinsic, tissue-type PLG activator,3’ and two types of intrinsic activator produced from circulating precursors” by one of two different pathways (see Fig I), one dependent upon and the other independent of the Hageman factor (blood coagulation factor XII).‘7.‘9+32The two intrinsic systems are both inhibited by C 1-inactivator. *’The intrinsic activator generated by the factor XII-independent pathway is immunologically related to, but chemically different from, urokinase.” The overall results of the present study indicate that oral administration of gliclazide in a dosage as great as 240 mg per day may influence some of the individual components of the extrinsic as well as the intrinsic pathways of fibrinolysis during the 6-month period of treatment. While fibrinolytic activities measured by the global assay of PLG activator in the NEF fluctuated insignificantly in the patients receiving 160 mg gliclazide, there was a small, but significant, fall in the patients receiving 240 mg (Fig 4A). This fall in activity seemed to be due to an artifact because it could be eliminated by addition to the euglobulin solutions of flufenamate (Fig 4B), indicating that a variable co-precipitation of inhibitors (different from the Cl-inactivator)
Fig 6. Median levels of fibrinolytic activity (mm’, logarithmic scale) in the DEF prepared from morning samples of blood (periods of sampling, abscissa) from type I diabetics given one daily dose of gliclazide of 160 mg (A) or 240 mg (B) for a period of 6 months. Closed signatures (A, 01 show the total activity of preoipiteble plasminogen activators. Open signatures (A, 0) depict the factor XII-dependent PLG ectivators. Levels of co-precipitated t-PA are too low to influence the determinations. Activities of the urokinase-related PLG activators in the samples ere represented by the differences (logarithmic) in activities between the closed and open signatures in A and B.
caused the measured fall in fibrinolytic activities of the euglobulins. This conclusion is supported by the finding that the specific determination of t-PA activity in euglobulins yielded stable values during the study (Table 1). In contrast, the baseline levels in plasma of t-PA antigen increased marginally during treatment with gliclazide (Table 1). These differences between activity and antigen determinations might be caused by nonenzymatic glycation of t-PA.24,25The measured increases of t-PA activity and t-PA antigen (Table 2) following venous occlusion was not significantly affected by gliclazide, indicating that endothelial release of t-PA upon stimulation remained unchanged. The precipitation by dextran sulphate of intrinsic activators raises the activity of the resulting DEFs to levels much higher than those obtained in the NEF.“,‘9 The DEF activities, the activities from the urokinaselike pathway, and the activities originating from the factor XII-dependent pathway of fibrinolysis showed negligible fluctuations during the seven periods of measurement (Fig 6). Factor XII and prekallikrein both are needed for the generation of intrinsic fibrinolytic activity by the factor XII-dependent pathway. Plasma levels of factor XII remained unchanged during the study, whereas prekallikrein increased significantly during administration of gliclazide (Fig 5). Some previous studies have reported observations indicating an increase in fibrinolytic activity during the treatment of maturity-onset diabetes (type II) with gliclazide.5.7 The study of type II diabetics reported by AlmCr” suffers from a lack of a wash-out interval, when the treatment was changed from chlorpropamide to gliclazide. A similar doubt applies to the study by Chan et aJ6 in which some of the patients had only a brief wash-out interval of approximately 2 weeks. Actually, in the latter study the improvements of the glycemic control seem to parallel the increased fibrinolytic activity. However, Paton et al” have reported that treatment of newly diagnosed, maturity-onset diabetics with gliclazide in doses varying from 80 mg to 320 mg per day did not raise the fibrinolytic activity (measured as the euglobulin clot lysis time) in parallel with an improved glycemic status. Our data obtained from type I diabetics cannot be compared directly with these previous studies. The design of our study was aimed at the procurement of data from a physiological model (type I diabetics treated with insulin) that was metabolically simpler than that represented by type II diabetics. The cause of the moderate increase in plasma concentrations of components involved in intrinsic (prekallikrein) and extrinsic (t-PA) fibrinolysis remains obscure. However, it might be suggested that the mechanisms are of differing natures.
942
GRAM, JESPERSEN, AND KDLD
Sulphonylureas have the potential to increase the plasma concentrations of liver enzymes (otherwise not involved in hemostasis), and an increasing production of prekallikrein by the liver would seem to cause increasing plasma concentrations following administration of gliclazide. The increase of plasma concentrations of t-PA antigen might be related to a more direct effect on the endothelial cells. Studies on endothelial cells and liver cells in culture would help to clarify these potential mechanisms. Another second-generation sulphonylurea, glipizide, was recently found to increase fivefold the synthesis and release of t-PA from cultured bovine aortic endothelial cells.34
Our study demonstrated that gliclazide possesses small, but significant, effects on variables involved in fibrinolysis, and that these effects are of an extrametabolic non-insulindependent nature. Whether the observed changes are of physiological significance remains to be determined in longterm prospective studies. ACKNOWLEDGMENT
Christian Schmidt is thanked for his interest. C-peptide was kindly determined by B. Reinholdt, Fredericia, Denmark. Thanks are due to Dr Tage Astrup, the South Jutland University Centre, Esbjerg, for valuable advice.
REFERENCES
1. Wilkens HJ, Back N: Fibrinolysis and risk factors of atherosclerotic disease, with special emphasis on diabetes mellitus. Circ Shock 51:125-143,1978 2. Fuller JH, Keen H, Jarrett RJ, et al: Haemostatic variables associated with diabetes and its complications. Br Med J 2:964-966, 1979 3. Haitas B, Cederholm-Williams SA, Moore J, et al: Abnormal endothelial release of fibrinolytic activity and fibronectin in diabetic microangiography. Diabetologia 27:493-496, 1984 4. Desnoyers P, Saint-Disier D: Gliclazide: Haemobiological properties. A synopsis with emphasis on inhibition of platelet coagulation factors. Royal Sot Med 20:19-27, 1980 5. AlmCr L-O: Vascular fibrinolytic activity in long term treatment with second generation sulfonylurea compounds. Acta Endocrinol239:53-55, 1980 (suppl) 6. Chan TK, Chan V, Teng CS, et al: Effects of gliclazide and glibenclamide on platelet function, fibrinolysis and metabolic control in diabetic patients with retinopathy. Semaine Hopitaux Paris 58:1197-1200, 1982 7. Almer L-O: Effect of chlorpropamide and gliclazide on plasminogen activator activity invascular walls in patients with maturity onset diabetes. Thromb Res 35:19-25, 1984 8. Jespersen J, Knudsen LH, Sidelmann J: The use of evacuated glass tubes for collection of blood samples for fibrinolytic assays. Thromb Res 25:173-176,1982 9. Poulsen JH, Jespersen J: A comparison of the determination of glycosylated hemoglobin by isoelectric focusing and cation-exchange chromatography on minicolumns. Stand J Clin Lab Invest 46:259263,1986 10. Kushner I: The phenomenon of the acute phase response. Ann NY Acad Sci 389:39-48,1982 11. Jespersen J, Sidelmann J: A study of the conditions and accuracy of the thrombin time assay of plasma fibrinogen. Acta Haematol67:2-7, 1982 12. Gram J, Jespersen J: A functional plasminogen assay utilizing the potentiating effect of fibrinogen to correct for the overestimation of plasminogen in pathological plasma samples. Thromb Haemost 53:255-259, 1985 13. Jespersen J, Astrup T: A study of the fibrin plate assay of fibrinolytic agents: Optimal conditions, reproducibility and precision. Haemostasis 13:301-315, 1983 14. Verheijen JH, Mullaart E, Chang GTG, et al: A simple, sensitive spectrophotometric assay for extrinsic (tissue-type) plasminogen activator applicable to measurements in plasma. Thromb Haemost 48:266-269. 1982 15. Gram J, Jespersen J: A simplified estimation of tissue plasminogen activator (t-PA) inhibition in human plasma. Fibrinolysis 1:33-37, 1987
16. Verheijen JH, Chang GTG, Kluft C: Evidence of the occurrence of a fast-acting inhibitor for tissue-type plasminogen activator in human plasma. Thromb Haemost 51:392-395, 1984 17. Kluft C, Wijngaards G, Jie AFH: Intrinsic plasma fibrinolysis: Involvement of urokinase related activity in the factor XIIindependent plasminogen proactivator pathway. J Lab Clin Med 103:408-419, 1984 18. Brakman P, Astrup T: Fibrinolytic activity in human blood assayed in a peptone-like system. Sangre 9:46-50, 1964 19. Astrup T, Rosa AT: A plasminogen-proactivator system in human blood effective in the absence of Hageman factor. Thromb Res 4:609-613, 1974 20. Kluft C: Studies on the fibrinolytic system in human plasma: Quantitative determination of plasminogen activators and proactivators. Thromb Haemost 41:365-383, 1979 21. Kluft C: Elimination of inhibition in euglobulin fibrinolysis by use of flufenamate: Involvement of C,-inactivator. Haemostasis 61351-369, 1977 22. Kluft C, Jie AFH: Underestimation of tissue-type plasminogen activator activity in plasma when euglobulin fractions are used for assessment. Fibrinolysis 2:147,1986 (suppl 1) 23. Jespersen J, Kluft C: Increased euglobulin fibrinolytic potential in women on oral contraceptives low in oestrogen. Levels of extrinsic and intrinsic plasminogen activators, prekallikrein, factor XII, and Cl-inactivator. Thromb Haemost 54:454-459,1985 24. Brownlee M, Vlassara H, Cerami A: Nonenzymatic glycosylation reduces the susceptibility of fibrin to degradation by plasmin. Diabetes 32:600-604, 1983 25. Geiger M, Binder BR: Plasminogen activation in diabetes mellitus. J Biol Chem 259:2676-2681, 1984 26. Fearnley GR, Vincent CT, Chakrabarti R: Reduction of blood fibrinolytic activity in diabetes mellitus by insulin. Lancet 2:1067, 1959 27. Epstein SE, Rosing DR, Brakman P, et al: Impaired fibrinolytic response to exercise in patients with type-IV hyperlipoproteinaemia. Lancet 2:631-634, 1970 28. Jespersen J, Gram J, Christensen JEJ: Neurophysiological variables and fibrinolysis in insulin-dependent diabetes treated with an aldose reductase inhibitor or placebo. A double-blind randomized study. Haemostasis 16:453-457, 1986 29. Jespersen J, Gram J, Bach E: A sequential study of plasma histidine-rich glycoprotein and plasminogen in patients with acute myocardial infarction. Thromb Haemost 51:99-102, 1984 30. Kluft C, Verheijen JH, Jie AFH, et al: The postoperative fibrinolytic shutdown: A rapidly reverting acute phase pattern for the fast-acting inhibitor of tissue-type plasminogen activator after trauma. Stand J Clin Lab Invest 45:605-610, 1985
FIBRINOLYSIS
AND ORAL ANTIDIABETIC
DRUGS
31. Rijken DC, Wijngaards G, Welbergen J: Relationship between tissue plasminogen activator and the activators in blood and vascular wall. Thromb Res 18:815-830, 1980 32. Jespersen J: The diurnal increase in euglobulin fibrinolytic activity in women using oral contraceptives and in normal women, and the generation of intrinsic tibrinolytic activity. Thromb Haemost 56:183-188, 1986
943
33. Paton RC, Kernoff PBA, Wales JK: Effects of diet and gliclazide on the haemostatic system of non-insulin-dependent diabetics. Br Med J 283:1018-1020, 1981 34. Kuo B-S, Korner G, Bjijrnsson TD: Effect of sulfonylureas on the synthesis and secretion of plasminogen activator from bovine aortic endothelial cells. Sixth International Washington Spring Symposium, Washington, DC, May 20-23, 1986 (abstr 222)